Package 'gDRplots'

Title: 'gDRplots'
Description: gDRplots provides static visualization functions and dedicated helpers for the gDR drug response platform. Designed for high-throughput screening (HTS), pharmacogenomics, and chemical genomics research, it enables comprehensive analysis of single-agent, co-dilution, and drug combination cell viability assays. The package supports quality control (plate layouts, control mapping), data normalization inspection, dose-response curves, distribution boxplots, drug synergy heatmaps with overlaid isobolograms, and combination index plots. Additionally, it integrates with PRISM and DepMap datasets to generate volcano plots and correlation scatter plots for molecular feature associations.
Authors: Marc Hafner [aut] (ORCID: <https://orcid.org/0000-0003-1337-7598>), Janina Smola [aut], Bartosz Czech [aut] (ORCID: <https://orcid.org/0000-0002-9908-3007>), Dariusz Scigocki [aut], Arkadiusz Gladki [aut, cre] (ORCID: <https://orcid.org/0000-0002-7059-6378>)
Maintainer: Arkadiusz Gladki <[email protected]>
License: Artistic-2.0
Version: 0.99.3
Built: 2026-10-01 13:42:58 UTC
Source: https://github.com/BiocStaging/gDRplots

Help Index

Analyze Chemical Genomics (CGS) Data and Perform GSEA

Description

This function analyzes chemical genomics (CGS) data, filters by compound mechanism of action (MOA), prepares the data for Gene Set Enrichment Analysis (GSEA), and performs GSEA for specified cell lines and metrics.

Usage

analyze_cgs(
  dt_metrics,
  metrics,
  cl_name = NULL,
  cellline1 = NULL,
  cellline2 = NULL,
  normalization_type = "RV"
)

Arguments

dt_metrics

A data.table containing screening data. Requires columns: 'drug_moa', 'DrugName', 'CellLineName', and columns specified in the 'metrics' argument.
metrics

A character vector specifying the response metrics to analyze: "x_mean", "x_AOC_range", "xc50", "ec50", "x_max".
cl_name

An optional string specifying a single cell line to analyze. If NULL (default), all cell lines in the data are analyzed. Should be NULL if cellline1 and cellline2 are provided.
cellline1

An optional string representing the first cell line name. Should be specified alongside cellline2.
cellline2

An optional string representing the second cell line name. Should be specified alongside cellline1.
normalization_type

A string with normalization types to be selected, one of: "GR" ("GRvalue") or "RV" ("RelativeViability"). Passed to prep_dt_response_metric_diff.

Value

A list of results, where each element corresponds to a cell line or cell line difference. Each result contains:

  • fgsea: GSEA results for the specified metrics,

  • metrics_diff: The prepped data.table used for the GSEA analysis,

  • moa_list: A list of DrugNames grouped by drug_moa.

Author(s)

Bartosz Czech [email protected]

Examples

dt_metrics <- qs2::qs_read(system.file("testdata/cgs_data.qs2", package = "gDRplots"))
analyze_cgs(dt_metrics, metrics = "xc50", cl_name = "CellLineName_1")
cellline1 <- "CellLineName_1"
cellline2 <- "CellLineName_2"
analyze_cgs(dt_metrics, "xc50", cl_name = NULL, cellline1, cellline2)

Compute distance between rows of a matrix.

Description

This function was lifted from the package factoextra and slightly adjusted. The default method was changed to "spearman" and an option was added to replace missing values in the resulting distance matrix with an arbitrary value. This option defaults to 0 so that the function can be called by iheatmapr::add_row_clustering and iheatmapr::add_col_clustering.

Usage

compute_distances(
  x,
  method = "spearman",
  use = "pairwise.complete.obs",
  stand = FALSE,
  dummy = 0,
  ...
)

Arguments

x

numeric matrix
method

character string specifying the distance measure to be used; must be on of: "euclidean", "maximum", "manhattan", "canberra", "binary", "minkowski", "pearson", "spearman" or "kendall"
use

character string specifying the method for computing covariances in the presence of missing values. Only used when "pearson", "spearman" or "kendall" chosen as a distance measure
stand

logical flag specifying whether the data should be standardized; if TRUE, columns are converted to z scores using scale
dummy

value to substitute for missing values; defaults to 0
...

arguments passed to internally called functions

Value

Object of class dist. NA and NaN are substituted by the value of dummy.

Control

The function offers different distance measures but since it is called internally, it will always use default arguments. As a result, should you want to do anything other than compute Spearman correlation distance and replace NAs with 0, you must edit this function definition accordingly. At the moment (December 2020) this is sufficient. If the control were to be given to the user, remove defaults for this definition and create a wrapper within plotly_metric_clustering.

The function was originally named computeDistances

Author(s)

Alboukadel Kassambara [email protected]

See Also

[factoextra::get_dist()]

Examples

x <- matrix(1:9, nrow = 3, ncol = 3)
rownames(x) <- letters[seq(NROW(x))]
compute_distances(x)

create a log-sequence

Description

Create a sequence of numbers growing in log-domain.

Usage

create_log_seq(start, end, length)

Arguments

start, end

numeric, lower and upper margins of the sequence
length

integer, resulting sequence length

Details

The result is a numeric vector of length length. Differences between items are constant in logarithmic domain and therefore geometrically increase in linear domain.

Value

A numeric vector, see Details.

Examples

create_log_seq(1, 2, 5)

Create a nested list of plots for PRISM data with combo metrics

Description

Create a nested list of plots for PRISM data with combo metrics

Usage

create_PRISM_plot_list_combo(
  drug1_name_vec,
  drug2_name_vec,
  dt_metrics,
  dt_scores = NULL,
  normalization_type_vec = "RV",
  metric = c("xc50", "x_mean", "x_max"),
  metric_scores = c("hsa_score", "bliss_score"),
  fit_source = "gDR",
  meta_data_path,
  feat_data_path,
  feature_sets,
  metadata_columns = NULL,
  clear_taxonomy_info = TRUE,
  with_decoding = FALSE
)

Arguments

drug1_name_vec

character vector with drug names to be plotted (identifiers DrugName)
drug2_name_vec

character vector with co-drug names to be plotted (identifiers DrugName_2)
dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and combo SummarizedExperiment
dt_scores

data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt(se, "scores") and combo SummarizedExperiment
normalization_type_vec

character vector with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability") or both
metric

character vector with names of metric; chosen from: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean")
metric_scores

character vector with names of combo metric; chosen from: "hsa_score"("Bliss Excess GR" or "Bliss Excess RV" - respectively depending on normalization_type), "bliss_score" ("Bliss Score GR" or "Bliss Score RV")
fit_source

string source name for metrics
meta_data_path

string path to metadata file describing all cancer models/cell lines which are referenced by a dataset contained within the DepMap portal. It is usually a file named Model.csv or Model.csv.gz.
feat_data_path

string path to the directory containing the molecular feature set file to load from DepMap.
feature_sets

character vector containing the names of the molecular feature sets to load from DepMap. These names should also correspond to the file names containing the feature data (without the extension, which is assumed to be csv or csv.gz)
metadata_columns

character vector with the metadata columns to load for DepMap cell lines
clear_taxonomy_info

logical flag whether to remove taxonomy information for gene names in table with the molecular feature sets from DepMap.
with_decoding

logical whether the feature OmicsArmLevelCNA, OmicsSomaticMutationsMatrixHotspot and OmicsSomaticMutationsMatrixDamaging should be encoded into a 0-1 scheme

Value

A named list with elements:

  • ls_plot nested list of plots for selected type of experiment

  • ls_assoc_data nested list of table with association data

Author(s)

Janina Smoła [email protected]

Examples

mae <- qs2::qs_read(system.file("testdata/finalMAE_combo_matrix_small.qs2",
                                 package = "gDRtestData"))
se_combo <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_combo, "Metrics")
drug1 <- unique(dt_metrics[[gDRutils::get_env_identifiers("drug_name")]])[1]
drug2 <- unique(dt_metrics[[gDRutils::get_env_identifiers("drug_name2")]])[1]
create_PRISM_plot_list_combo(
  drug1_name_vec = drug1,
  drug2_name_vec = drug2,
  dt_metrics = dt_metrics,
  meta_data_path = system.file("depmap_data/Model.csv.gz", package = "gDRtestData"),
  feat_data_path = system.file("depmap_data", package = "gDRtestData"),
  feature_sets = c("CRISPRGeneEffect")
)

Create a nested list of plots for PRISM data with single-agent metrics

Description

Create a nested list of plots for PRISM data with single-agent metrics

Usage

create_PRISM_plot_list_sa(
  drug_name_vec,
  dt_metrics,
  dt_average = NULL,
  normalization_type_vec = "RV",
  metric = c("xc50", "x_mean", "x_max"),
  fit_source = "gDR",
  meta_data_path,
  feat_data_path,
  feature_sets,
  metadata_columns = NULL,
  clear_taxonomy_info = TRUE,
  with_decoding = FALSE
)

Arguments

drug_name_vec

character vector with drug names to be plotted (identifiers DrugName)
dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and single-agent SummarizedExperiment
normalization_type_vec

character vector with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability") or both
metric

character vector with names of metric; chosen from: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean")
fit_source

string source name for metrics
meta_data_path

string path to metadata file describing all cancer models/cell lines which are referenced by a dataset contained within the DepMap portal. It is usually a file named Model.csv or Model.csv.gz.
feat_data_path

string path to the directory containing the molecular feature set file to load from DepMap.
feature_sets

character vector containing the names of the molecular feature sets to load from DepMap. These names should also correspond to the file names containing the feature data (without the extension, which is assumed to be csv or csv.gz)
metadata_columns

character vector with the metadata columns to load for DepMap cell lines
clear_taxonomy_info

logical flag whether to remove taxonomy information for gene names in table with the molecular feature sets from DepMap.
with_decoding

logical whether the feature OmicsArmLevelCNA, OmicsSomaticMutationsMatrixHotspot and OmicsSomaticMutationsMatrixDamaging should be encoded into a 0-1 scheme

Value

A named list with elements:

  • ls_plot nested list of plots for selected type of experiment

  • ls_assoc_data nested list of table with association data

Author(s)

Janina Smoła [email protected]

Examples

mae <- qs2::qs_read(system.file("testdata/finalMAE_prism.qs2", package = "gDRtestData"))
se <- mae[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
drug_name <- unique(dt_metrics[[gDRutils::get_env_identifiers("drug_name")]])[1]
create_PRISM_plot_list_sa(
  drug_name_vec = drug_name,
  dt_metrics = dt_metrics,
  meta_data_path = system.file("depmap_data/Model.csv.gz", package = "gDRtestData"),
  feat_data_path = system.file("depmap_data", package = "gDRtestData"),
  feature_sets = c("CRISPRGeneEffect")
)

Create a list of PRISM association table

Description

Create a list of PRISM association table

Usage

create_PRISM_summary_list(assoc_summary_RV, assoc_summary_GR = NULL)

Arguments

assoc_summary_RV

A data.table with associations for normalization type of "Relative Viability", outputted by gDRplots::prep_assoc_summary()
assoc_summary_GR

A data.table with associations for normalization type of "GR Value", outputted by gDRplots::prep_assoc_summary()

Value

A list of table split by drug name and normalization type

Author(s)

Janina Smoła [email protected]

Examples

dir_path <- system.file("testdata", package = "gDRplots")
ls_file <- list.files(dir_path, pattern = "[.]xlsx$")
assoc_tab <- prep_assoc_summary(dir_path = dir_path, ls_file = ls_file)
create_PRISM_summary_list(
  assoc_summary_RV = assoc_tab[grepl("RV", assoc_tab[["src"]]), ],
  assoc_summary_GR = assoc_tab[grepl("GR", assoc_tab[["src"]]), ]
)

Fill missing values in the color map for annotation

Description

Fill missing values in the color map for annotation

Usage

fill_ann_color_map(dt_ann, map_ann)

Arguments

dt_ann

data.table with the annotations
map_ann

list with the annotations

Value

list with color mapping for the annotations with missing items filled in

Author(s)

Janina Smoła [email protected]

See Also

pheatmap_with_anno_sa pheatmap_with_anno_combo

Examples

annotation_manual <- data.table::data.table(
  CellLineName = c("cellline_AA", "cellline_EA", "cellline_IB", "cellline_MC", "cellline_BC"),
  mut_A = c(0, 0, 1, 2, 3),
  mut_B = c("yes", "yes", "no", NA, NA),
  mut_C = c("AA", "AA", "AB", NA, "B")
)

annotation_map <- list(
  mut_A = c("1" = "coral", "0" = "cadetblue"),
  mut_B = c("yes" = "black", "no" = "grey90", "not_checked" = "lightblue"),
  mut_C = c("AA" = "red", "AB" = "orange", "B" = "yellow")
)

fill_ann_color_map(dt_ann = annotation_manual, map_ann = annotation_map)

Create color map for annotation

Description

Create color map for annotation

Usage

get_ann_color_map(dt_ann)

Arguments

dt_ann

data.table with the annotations

Value

list with color mapping for the annotations

Author(s)

Janina Smoła [email protected]

See Also

pheatmap_qc

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]][2:5, ]
dt_average <- gDRutils::convert_se_assay_to_dt(se = se, assay_name = "Averaged")
dt_ann <- dt_average[,.SD, .SDcols = c("Tissue", "ReferenceDivisionTime")]

get_ann_color_map(dt_ann)

Calculate the luminance of a color

Description

Calculate the luminance of a color

Usage

get_col_luminance(col_name)

Arguments

col_name

string representing a valid color

Value

single element numeric vector

Examples

get_col_luminance("blue")
get_col_luminance("red")
get_col_luminance("#000000")
get_col_luminance("#906090")
get_col_luminance("#906090F2")

Get Legend Title

Description

Get Legend Title

Usage

get_hm_title(metric = "xc50", normalization_type = "GR", dataset_name = NULL)

Arguments

metric

string name of the metric one of: "xc50"("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or x_mean" ("GR Mean" or "RV Mean")
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
dataset_name

string name of dataset

Value

character title for heatmap

Examples

get_hm_title(dataset_name = "Dateset DX123",
             metric = "x_mean",
             normalization_type = "GR")

get_hm_title(metric = "xc50",
             normalization_type = "GR")

get_iso_colors

Description

get_iso_colors

Usage

get_iso_colors(normalization_type = c("RV", "GR"))

Arguments

normalization_type

charvec normalization_types expected in the data

Value

named charvec with iso colors

Examples

get_iso_colors()

Create list of qualitative colors

Description

Create list of qualitative colors

Usage

get_qual_colors(n = NULL)

Arguments

n

number of required colors

Value

vector with hex colors from qualitative palettes

Author(s)

Janina Smoła [email protected]

Examples

get_qual_colors()
get_qual_colors(0)
get_qual_colors(5)
get_qual_colors(35)

Plot heatmaps of fitted values for combination metrics data

Description

Plot heatmaps of fitted values for combination metrics data

Usage

heatmap_combo_metrics(
  dt_excess,
  dt_isobolograms = NULL,
  drug1_name,
  drug2_name,
  cl_name,
  metric = "smooth",
  normalization_type = "GR",
  iso_levels = c("0.25", "0.5", "0.75"),
  colors_vec_smooth = NULL,
  colors_vec_excess = NULL,
  limit = NULL,
  no_breaks = 50,
  swap_axes = FALSE,
  show_values = FALSE
)

Arguments

dt_excess

data.table representing data from the excess assay, outputted by gDRutils::convert_se_assay_to_dt(se, "excess") and combo SummarizedExperiment
dt_isobolograms

data.table representing data from the isobolograms assay, outputted by gDRutils::convert_se_assay_to_dt(se, "isobolograms") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_name

string with cell line to be plotted (identifiers CellLineName)
metric

string name of combo metric; one of: "smooth"("Smooth GR" or "Smooth RV" - respectively depending on normalization_type), "hsa_excess"("Bliss Excess GR" or "Bliss Excess RV") or "bliss_excess" ("Bliss Excess GR" or "Bliss Excess RV")
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
iso_levels

character vector with isobologram levels to be selected
colors_vec_smooth

character vector of colors (valid names or hex codes) used in the heatmap for smooth values; the default is the dark purple-light grey palette
colors_vec_excess

character vector of colors (valid name or hex codes) used in the heatmap for excess values; the default is the blue-light grey-red color scale
limit

numeric vector of length two providing limits of the scale. Use NA to refer to the existing minimum or maximum
no_breaks

numeric number of breaks on scale
swap_axes

logical flag indicating whether to swap the axes with drugs of the heatmap
show_values

logical flag indicating whether to show values of the metric on the heatmap

Value

ggplot object containing heatmap with the values for the selected combo metric for selected drugs and cell line with selected isoline (when chosen)

Examples

cl_name <- "cellline_BC"
drug1_name <- "drug_001"
drug2_name <- "drug_026"

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_excess <- gDRutils::convert_se_assay_to_dt(se, "excess")
dt_isobolograms <- gDRutils::convert_se_assay_to_dt(se, "isobolograms")

heatmap_combo_metrics(dt_excess,
                      dt_isobolograms,
                      drug1_name, drug2_name,
                      cl_name,
                      normalization_type = "GR")

heatmap_combo_metrics(dt_excess,
                      dt_isobolograms,
                      drug1_name, drug2_name,
                      cl_name,
                      normalization_type = "GR",
                      swap_axes = TRUE)

heatmap_combo_metrics(dt_excess,
                      dt_isobolograms,
                      drug1_name, drug2_name,
                      cl_name,
                      metric = "hsa_excess",
                      normalization_type = "GR",
                      limit = c(NA, 0.1))

cl_name <- "cellline_JE"
drug1_name <- "drug_011"
drug2_name <- "drug_026"

heatmap_combo_metrics(dt_excess,
                      dt_isobolograms,
                      drug1_name, drug2_name,
                      cl_name,
                      metric = "bliss_excess",
                      normalization_type = "RV",
                      iso_levels = "0.5")

heatmap_combo_metrics(dt_excess,
                      dt_isobolograms,
                      drug1_name, drug2_name,
                      cl_name,
                      metric = "bliss_excess",
                      normalization_type = "RV",
                      iso_levels = NULL)

Plot panel of heatmaps of fitted values for combination metrics data

Description

Plot panel of heatmaps of fitted values for combination metrics data

Usage

heatmap_combo_metrics_panel(
  dt_excess,
  dt_isobolograms = NULL,
  drug1_name,
  drug2_name,
  cl_name,
  normalization_type = "GR",
  iso_levels = c("0.25", "0.5", "0.75"),
  colors_vec_smooth = NULL,
  colors_vec_excess = NULL,
  no_breaks = 50,
  as_list = FALSE,
  one_row_panel = FALSE,
  swap_axes = FALSE,
  show_values = FALSE
)

Arguments

dt_excess

data.table representing data from the excess assay, outputted by gDRutils::convert_se_assay_to_dt(se, "excess") and combo SummarizedExperiment
dt_isobolograms

data.table representing data from the isobolograms assay, outputted by gDRutils::convert_se_assay_to_dt(se, "isobolograms") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_name

string with cell line to be plotted (identifiers CellLineName)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
iso_levels

character vector with isobologram levels to be selected
colors_vec_smooth

character vector of colors (valid names or hex codes) used in the heatmap for smooth values; the default is the dark purple-light grey palette
colors_vec_excess

character vector of colors (valid name or hex codes) used in the heatmap for excess values; the default is the blue-light grey-red color scale
no_breaks

numeric number of breaks on scale
as_list

logical flag whether return list of plot or panel
one_row_panel

logical flag whether return panel 2x2 (containing heatmaps for combination metrics and CI plot) or 3x1 (containing only heatmaps for combination metrics); it is working only for as_list = TRUE
swap_axes

logical flag indicating whether to swap the axes with drugs of the heatmap
show_values

logical flag indicating whether to show values of the metric on the heatmap

Value

ggplot object containing panel with heatmaps with value for excess assays for selected drugs and cell line with selected isoline and comparison of iso levels or list of ggplot object containing these plots.

Examples

cl_name <- "cellline_BC"
drug1_name <- "drug_001"
drug2_name <- "drug_026"

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_excess <- gDRutils::convert_se_assay_to_dt(se, "excess")
dt_isobolograms <- gDRutils::convert_se_assay_to_dt(se, "isobolograms")

heatmap_combo_metrics_panel(dt_excess,
                            dt_isobolograms,
                            drug1_name, drug2_name,
                            cl_name,
                            normalization_type = "GR")

cl_name <- "cellline_JE"
drug1_name <- "drug_011"
drug2_name <- "drug_026"

heatmap_combo_metrics_panel(dt_excess,
                            dt_isobolograms,
                            drug1_name, drug2_name,
                            cl_name,
                            normalization_type = "RV",
                            iso_levels = "0.5",
                            as_list = TRUE)

heatmap_combo_metrics_panel(dt_excess,
                            dt_isobolograms,
                            drug1_name, drug2_name,
                            cl_name,
                            normalization_type = "RV",
                            iso_levels = "0.5",
                            one_row_panel = TRUE)

heatmap_combo_metrics_panel(dt_excess,
                            dt_isobolograms,
                            drug1_name, drug2_name,
                            cl_name,
                            normalization_type = "RV",
                            iso_levels = NULL,
                            as_list = FALSE,
                            swap_axes = FALSE)

heatmap_combo_metrics_panel(dt_excess,
                            dt_isobolograms,
                            drug1_name, drug2_name,
                            cl_name,
                            normalization_type = "RV",
                            iso_levels = NULL,
                            as_list = FALSE,
                            swap_axes = TRUE)

Plot heatmaps of averaged values for combination data

Description

Plot heatmaps of averaged values for combination data

Usage

heatmap_combo_with_isoref(
  dt_excess,
  dt_isobolograms,
  drug1_name,
  drug2_name,
  cl_name,
  normalization_type = "GR",
  metric = "smooth",
  iso_levels = "0.5",
  colors_vec = NULL,
  colors_vec_iso = NULL,
  no_breaks = 50,
  swap_axes = FALSE
)

Arguments

dt_excess

data.table representing data from the excess assay, outputted by gDRutils::convert_se_assay_to_dt(se, "excess") and combo SummarizedExperiment
dt_isobolograms

data.table representing data from the isobolograms assay, outputted by gDRutils::convert_se_assay_to_dt(se, "isobolograms") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_name

string with cell line to be plotted (identifiers CellLineName)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "smooth" ("Smooth GR" or "Smooth RV" - respectively depending on normalization_type) "hsa_excess" ("Bliss Excess GR" or "Bliss Excess RV") or "bliss_excess" ("Bliss Excess GR" or "Bliss Excess RV")
iso_levels

character vector with isobologram levels to be selected; when NULL - no isolines will be displayed
colors_vec

character vector of colors (valid name or hex) used in heatmap; the default depends on metric: for "smooth" - the dark purple-light grey palette and for "hsa_excess" and "bliss_excess" - the blue-light grey-red color scale
colors_vec_iso

character vector of colors (valid name or hex) used for the isolines; the default is the dark red-orange palette
no_breaks

numeric number of breaks on scale
swap_axes

logical flag whether to swap the axes with drugs of the heatmap

Value

ggplot object containing heatmap for fitted values and reference data for isobolograms for selected drug and co-drug and selected cell line

Examples

cl_name <- "cellline_BC"
drug1_name <- "drug_001"
drug2_name <- "drug_026"

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_excess <- gDRutils::convert_se_assay_to_dt(se, "excess")
dt_isobolograms <- gDRutils::convert_se_assay_to_dt(se, "isobolograms")

heatmap_combo_with_isoref(dt_excess,
                          dt_isobolograms,
                          drug1_name, drug2_name,
                          cl_name)

heatmap_combo_with_isoref(dt_excess,
                          dt_isobolograms,
                          drug1_name, drug2_name,
                          cl_name,
                          metric = "hsa_excess",
                          iso_levels = c("-0.2", "0.2"))

heatmap_combo_with_isoref(dt_excess,
                          dt_isobolograms,
                          drug1_name, drug2_name,
                          cl_name,
                          iso_levels = NULL,
                          colors_vec = c("darkcyan", "snow", "darkorange"))

heatmap_combo_with_isoref(dt_excess,
                          dt_isobolograms,
                          drug1_name, drug2_name,
                          cl_name,
                          normalization_type = "RV",
                          iso_levels = c("0.25", "0.75"),
                          swap_axes = FALSE)

heatmap_combo_with_isoref(dt_excess,
                          dt_isobolograms,
                          drug1_name, drug2_name,
                          cl_name,
                          normalization_type = "RV",
                          iso_levels = c("0.25", "0.75"),
                          swap_axes = TRUE)

heatmap_combo_with_isoref(dt_excess,
                          dt_isobolograms,
                          drug1_name, drug2_name,
                          cl_name,
                          metric = "hsa_excess",
                          iso_levels = c("0.25", "0.75"),
                          colors_vec_iso = c("0.25" = "darkcyan",
                                             "0.75" = "darkblue"),
                          swap_axes = FALSE)

Plot panel of heatmaps with fitted and reference data for isobolograms

Description

Plot panel of heatmaps with fitted and reference data for isobolograms

Usage

heatmap_combo_with_isoref_panel(
  dt_excess,
  dt_isobolograms,
  drug1_name,
  drug2_name,
  cl_names,
  normalization_type = "GR",
  metric = "smooth",
  iso_levels = "0.5",
  colors_vec = NULL,
  no_breaks = 50,
  swap_axes = FALSE
)

Arguments

dt_excess

data.table representing data from the excess assay, outputted by gDRutils::convert_se_assay_to_dt(se, "excess") and combo SummarizedExperiment
dt_isobolograms

data.table representing data from the isobolograms assay, outputted by gDRutils::convert_se_assay_to_dt(se, "isobolograms") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_names

character vector with cell line names to be plotted (Cell Line Name); if NULL - all available cell lines will be plotted
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "smooth" ("Smooth GR" or "Smooth RV" - respectively depending on normalization_type) "hsa_excess" ("Bliss Excess GR" or "Bliss Excess RV") or "bliss_excess" ("Bliss Excess GR" or "Bliss Excess RV")
iso_levels

character vector with isobologram levels to be selected; when NULL - no isolines will be displayed
colors_vec

character vector of colors (valid name or hex) used in heatmap; the default depends on metric: for "smooth" - the dark purple-light grey palette and for "hsa_excess" and "bliss_excess" - the blue-light grey-red color scale
no_breaks

numeric number of breaks on scale
swap_axes

logical flag whether to swap the axes with drugs of the heatmap

Value

ggplot object containing panel with heatmaps for fitted values and reference data for isobolograms for selected drug and co-drug and all selected cell line

Examples

cl_names <-
  c("cellline_AA", "cellline_EA", "cellline_IB",
  "cellline_MC", "cellline_BC", "cellline_FD")

drug1_name <- "drug_001"
drug2_name <- "drug_026"

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_excess <- gDRutils::convert_se_assay_to_dt(se, "excess")
dt_isobolograms <- gDRutils::convert_se_assay_to_dt(se, "isobolograms")

heatmap_combo_with_isoref_panel(dt_excess,
                                dt_isobolograms,
                                drug1_name, drug2_name,
                                cl_names)

dt_excess_2 <- data.table::copy(dt_excess)
invisible(dt_excess_2[CellLineName %in% cl_names[1:2],
                      Concentration := Concentration / 10])

heatmap_combo_with_isoref_panel(dt_excess_2,
                                dt_isobolograms,
                                drug1_name, drug2_name,
                                cl_names,
                                iso_levels = c("0.25", "0.75"),
                                colors_vec = c("darkcyan", "snow", "coral"))

#' Plot panel of heatmaps with fitted and reference data for isobolograms

Description

#' Plot panel of heatmaps with fitted and reference data for isobolograms

Usage

heatmap_combo_with_isoref_panel_common(
  dt_excess,
  dt_isobolograms,
  drug1_name,
  drug2_name,
  cl_names,
  normalization_type = "GR",
  metric = "smooth",
  iso_levels = "0.5",
  colors_vec = NULL,
  no_breaks = 50,
  swap_axes = FALSE
)

Arguments

dt_excess

data.table representing data from the excess assay, outputted by gDRutils::convert_se_assay_to_dt(se, "excess") and combo SummarizedExperiment
dt_isobolograms

data.table representing data from the isobolograms assay, outputted by gDRutils::convert_se_assay_to_dt(se, "isobolograms") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_names

character vector with cell line names to be plotted (Cell Line Name); if NULL - all available cell lines will be plotted
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "smooth" ("Smooth GR" or "Smooth RV" - respectively depending on normalization_type) "hsa_excess" ("Bliss Excess GR" or "Bliss Excess RV") or "bliss_excess" ("Bliss Excess GR" or "Bliss Excess RV")
iso_levels

character vector with isobologram levels to be selected; when NULL - no isolines will be displayed
colors_vec

character vector of colors (valid name or hex) used in heatmap; the default depends on metric: for "smooth" - the dark purple-light grey palette and for "hsa_excess" and "bliss_excess" - the blue-light grey-red color scale
no_breaks

numeric number of breaks on scale
swap_axes

logical flag whether to swap the axes with drugs of the heatmap

Value

ggplot object containing panel with heatmaps for fitted values and reference data for isobolograms for selected drug and co-drug by list of cell lines

Examples

cl_names <-
  c("cellline_AA", "cellline_EA", "cellline_IB",
  "cellline_MC", "cellline_BC", "cellline_FD")

drug1_name <- "drug_001"
drug2_name <- "drug_026"

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_excess <- gDRutils::convert_se_assay_to_dt(se, "excess")
dt_isobolograms <- gDRutils::convert_se_assay_to_dt(se, "isobolograms")

heatmap_combo_with_isoref_panel_common(dt_excess,
                                       dt_isobolograms,
                                       drug1_name, drug2_name,
                                       cl_names)

heatmap_combo_with_isoref_panel_common(dt_excess,
                                       dt_isobolograms,
                                       drug1_name, drug2_name,
                                       cl_names,
                                       iso_levels = c("0.25", "0.5"))

heatmap_combo_with_isoref_panel_common(dt_excess,
                                       dt_isobolograms,
                                       drug1_name, drug2_name,
                                       cl_names,
                                       metric = "hsa_excess",
                                       iso_levels = c("-0.25", "0.25"))

heatmap_combo_with_isoref_panel_common(dt_excess,
                                       dt_isobolograms,
                                       drug1_name, drug2_name,
                                       cl_names,
                                       normalization_type = "RV",
                                       iso_levels = NULL,
                                       colors_vec = c("darkcyan", "snow", "darkorange"),
                                       swap_axes = FALSE)

heatmap_combo_with_isoref_panel_common(dt_excess,
                                       dt_isobolograms,
                                       drug1_name, drug2_name,
                                       cl_names,
                                       normalization_type = "RV",
                                       iso_levels = NULL,
                                       colors_vec = c("darkcyan", "snow", "darkorange"),
                                       swap_axes = TRUE)

heatmap_combo_with_isoref_panel_common(dt_excess,
                                       dt_isobolograms,
                                       drug1_name, drug2_name,
                                       cl_names,
                                       metric = "hsa_excess",
                                       iso_levels = NULL,
                                       swap_axes = FALSE)

Plot panel of heatmaps with fitted and reference data for isobolograms

Description

This function is dedicated to cases in which given cell lines are exposed to drugs of different concentrations and have almost no or no common values.

Usage

heatmap_combo_with_isoref_panel_independent(
  dt_excess,
  dt_isobolograms,
  drug1_name,
  drug2_name,
  cl_names,
  normalization_type = "GR",
  metric = "smooth",
  iso_levels = "0.5",
  colors_vec = NULL,
  no_breaks = 50,
  swap_axes = FALSE
)

Arguments

dt_excess

data.table representing data from the excess assay, outputted by gDRutils::convert_se_assay_to_dt(se, "excess") and combo SummarizedExperiment
dt_isobolograms

data.table representing data from the isobolograms assay, outputted by gDRutils::convert_se_assay_to_dt(se, "isobolograms") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_names

character vector with cell line names to be plotted (Cell Line Name); if NULL - all available cell lines will be plotted
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "smooth" ("Smooth GR" or "Smooth RV" - respectively depending on normalization_type) "hsa_excess" ("Bliss Excess GR" or "Bliss Excess RV") or "bliss_excess" ("Bliss Excess GR" or "Bliss Excess RV")
iso_levels

character vector with isobologram levels to be selected; when NULL - no isolines will be displayed
colors_vec

character vector of colors (valid name or hex) used in heatmap; the default depends on metric: for "smooth" - the dark purple-light grey palette and for "hsa_excess" and "bliss_excess" - the blue-light grey-red color scale
no_breaks

numeric number of breaks on scale
swap_axes

logical flag whether to swap the axes with drugs of the heatmap

Value

ggplot object containing panel with heatmaps for fitted values and reference data for isobolograms for selected drug and co-drug by list of cell lines

Examples

cl_names <-
  c("cellline_AA", "cellline_EA", "cellline_IB",
  "cellline_MC", "cellline_BC", "cellline_FD")

drug1_name <- "drug_001"
drug2_name <- "drug_026"

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_excess <- gDRutils::convert_se_assay_to_dt(se, "excess")
dt_isobolograms <- gDRutils::convert_se_assay_to_dt(se, "isobolograms")

heatmap_combo_with_isoref_panel_independent(dt_excess,
                                            dt_isobolograms,
                                            drug1_name, drug2_name,
                                            cl_names)

heatmap_combo_with_isoref_panel_independent(dt_excess,
                                            dt_isobolograms,
                                            drug1_name, drug2_name,
                                            cl_names,
                                            iso_levels = c("-0.25", "0.25"))

heatmap_combo_with_isoref_panel_independent(
  dt_excess,
  dt_isobolograms,
  drug1_name, drug2_name,
  cl_names = c("cellline_FD", "cellline_MC", "cellline_AA", "cellline_EA"),
  iso_levels =  c("-0.25", "-0.05", "0.2", "0.65"))

heatmap_combo_with_isoref_panel_independent(dt_excess,
                                            dt_isobolograms,
                                            drug1_name, drug2_name,
                                            cl_names,
                                            normalization_type = "RV",
                                            iso_levels = NULL,
                                            colors_vec = c("darkcyan", "snow", "darkorange"),
                                            swap_axes = TRUE)

Plot heatmap of mapping controls to treated for single-agent and combo data to control quality of the data

Description

Plot heatmap of mapping controls to treated for single-agent and combo data to control quality of the data

Usage

heatmap_control_mapping_qc(dt_treat, dt_controls)

Arguments

dt_treat

data.table representation of the data in RawTreated assay, outputted by gDRutils::convert_se_assay_to_dt(se, "RawTreated") and SummarizedExperiment with chosen data type: single-agent or combo
dt_controls

data.table representation of the data in Controls assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Controls") and SummarizedExperiment with chosen data type: single-agent or combo

Value

pheatmap object containing hetamap of mapping controls to treated

Author(s)

Bartosz Czech [email protected]

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_treat <- gDRutils::convert_se_assay_to_dt(se, "RawTreated")
dt_controls <- gDRutils::convert_se_assay_to_dt(se, "Controls")

heatmap_control_mapping_qc(dt_treat = dt_treat,
                           dt_controls = dt_controls)

heatmap_control_mapping_qc(dt_treat = dt_treat[1:1350, ],
                           dt_controls = dt_controls)

dt_treat_NA <- dt_treat[-c(1:135, 270:405),]
heatmap_control_mapping_qc(dt_treat = dt_treat_NA,
                           dt_controls = dt_controls)

dt_controls_NA <- dt_controls[-c(1:305, 611:763, 1221:1750),]
heatmap_control_mapping_qc(dt_treat = dt_treat,
                           dt_controls = dt_controls_NA)

heatmap_control_mapping_qc(dt_treat = dt_treat,
                           dt_controls = dt_controls[1:3660, ])

Determine whether or not a color is dark

Description

Determine whether or not a color is dark

Usage

is_color_dark(col_name)

Arguments

col_name

string representing a valid color

Value

logical flag

Examples

is_color_dark("blue")
is_color_dark("red")
is_color_dark("#000000")

Determine whether or not a color name is valid

Description

A name of color is valid when either is a color name listed by grDevices::colors or a hexadecimal string of the form #rrggbb

Usage

is_valid_color(col_name)

Arguments

col_name

string representing a valid color

Value

logical flag

Examples

is_valid_color("darkblue")
is_valid_color("#FF8C00")
is_valid_color("#FF8C00DC")
is_valid_color("RED")

Plot pretty heatmap for single-agent or combo data to control quality of the data

Description

Plot pretty heatmap for single-agent or combo data to control quality of the data

Usage

pheatmap_qc(
  dt_average,
  normalization_type = "GR",
  metric = "x",
  fit_source = "gDR",
  hm_title = NA,
  colors_vec = c("black", "grey99"),
  no_breaks = 50,
  cluster_rows = TRUE,
  distfun = compute_distances,
  lbl_by_CellLineName = FALSE,
  lbl_by_DrugName = FALSE
)

Arguments

dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt and SummarizedExperiment with chosen data type: single-agent or combo
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "x" (value of "GR" or "RV" itself - respectively depending on normalization_type), or "x_std" (standard deviation)
fit_source

string source name for metrics
hm_title

string plot title
colors_vec

character vector of colors (valid name or hex) used in heatmap note that for metric "x" the first color will be assigned to the min value, and the last one - to the max; for "x_std" - that will be reversed
no_breaks

numeric number of breaks on scale used for mapping values to colors
cluster_rows

logical flag whether rows should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
distfun

function used to compute the distance (dissimilarity) between rows; used for the dendrogram when cluster_rows is set to TRUE; the default is compute_distances using Spearman method.
lbl_by_CellLineName

logical flag whether heatmap should be described by CellLineNames instead of clid
lbl_by_DrugName

logical flag whether heatmap should be described by DrugName instead of Gnumber

Value

pheatmap object containing heatmap for selected metric with annotation - if given

See Also

pheatmap::pheatmap

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]][2:5, ]
dt_average <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Averaged")

hm_1 <- pheatmap_qc(dt_average = dt_average)

hm_2 <- pheatmap_qc(dt_average = dt_average,
                    normalization_type = "RV",
                    colors_vec = c("darkblue", "grey90"),
                    lbl_by_CellLineName = TRUE,
                    lbl_by_DrugName = TRUE)

patchwork::wrap_elements(hm_1[["gtable"]])
patchwork::wrap_elements(hm_2[["gtable"]])

se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_average <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Averaged")
hm_3 <- pheatmap_qc(dt_average = dt_average,
                    cluster_rows = FALSE)
hm_4 <- pheatmap_qc(dt_average = dt_average,
                    metric = "x_std",
                    cluster_rows = FALSE)

patchwork::wrap_elements(hm_3[["gtable"]])
patchwork::wrap_elements(hm_4[["gtable"]])

Plot pretty heatmap with annotations for co-dilution data

Description

Plot pretty heatmap with annotations for co-dilution data

Usage

pheatmap_with_anno_cd(
  dt_metrics,
  normalization_type = "GR",
  metric = "xc50",
  fit_source = "gDR",
  hm_title = NA,
  colors_vec = NULL,
  no_breaks = 50,
  cluster_rows = TRUE,
  cluster_cols = TRUE,
  distfun = compute_distances,
  annotation_row = NULL,
  annotation_col = NULL,
  annotation_colors = NULL
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt and co-dilution SummarizedExperiment
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean")
fit_source

string source name for metrics
hm_title

string plot title
colors_vec

character vector of colors (valid name or hex) used in heatmap; note that the first color will be assigned to the min value, and the last one - to the max
no_breaks

numeric number of breaks on scale used for mapping values to colors
cluster_rows

logical flag whether indicating rows should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
cluster_cols

logical flag indicating whether columns should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
distfun

function used to compute the distance (dissimilarity) between both rows and columns; used for the dendrogram when cluster_rows or cluster_cols is set to TRUE the default is compute_distances using Spearman method.
annotation_row

data.table that specifies the annotations shown on the left side of the heatmap. Each row defines the features for a specific row. The rows in the data and in the annotation are matched using corresponding names from the required DrugName, DrugName_2 and Concentration_2 columns. Note that color schemes takes into account if the variable is continuous or discrete.
annotation_col

data.table that specifies the annotations shown above the heatmap. Each row defines the features for a specific column. The columns in the data and in the annotation are matched using corresponding names from the required CellLineName column. Note that color schemes takes into account if the variable is continuous or discrete.
annotation_colors

named list for specifying annotation_col and annotation_row track colors manually; note list is named with annotation name (column names of annotation_row - without DrugName and column names of annotation_col - without CellLineName), each list item is named vector with valid color name for each value described in annotation_row and in annotation_col, respectively. Not described elements will be colored by default.

Value

A named list with elements:

  • data a list containing the information visualized in the heatmap:

    • matrix data shown in the heatmap for the selected metric.

    • annotation_row a table with row annotations (for DrugName), if provided.

    • annotation_col a table with column annotations (for CellLineName), if provided.

  • pheatmap object containing the heatmap itself.

Author(s)

Janina Smoła [email protected]

See Also

pheatmap::pheatmap

Examples

mae <- gDRutils::get_synthetic_data("combo_codilution_small")
se <- mae[[gDRutils::get_supported_experiments("cd")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Metrics")

output <- pheatmap_with_anno_cd(dt_metrics = dt_metrics)
hm_1 <- output[["heatmap"]]
patchwork::wrap_elements(hm_1[["gtable"]])

annotation_manual_col <-
  unique(dt_metrics[, c("CellLineName", "Tissue"), with = FALSE])
annotation_manual_row <-
  unique(dt_metrics[, c("DrugName", "DrugName_2", "Concentration_2",
                        "drug_moa", "drug_moa_2"),
                    with = FALSE])
annotation_map <-
  get_ann_color_map(unique(dt_metrics[, c("Tissue", "drug_moa", "drug_moa_2"), with = FALSE]))

output <- pheatmap_with_anno_cd(dt_metrics = dt_metrics,
                                normalization_type = "RV",
                                metric = "x_mean",
                                colors_vec = c("darkblue", "grey90"),
                                annotation_row = annotation_manual_row,
                                annotation_col = annotation_manual_col,
                                annotation_colors = annotation_map)
hm_2 <- output[["heatmap"]]
patchwork::wrap_elements(hm_2[["gtable"]])

annotation_manual <- data.table::data.table(
  CellLineName =
    c("cellline_AA", "cellline_EA", "cellline_IB", "cellline_MC", "cellline_BC"),
  mut_A = c(1, 1, 1, 0, 0),
  mut_B = c("yes", "yes", "no", "no", "no")
)
annotation_map <- list(
  mut_A = c("1" = "coral", "0" = "cadetblue"),
  mut_B = c("yes" = "black", "no" = "grey90")
)

output <- pheatmap_with_anno_cd(dt_metrics = dt_metrics,
                                annotation_col = annotation_manual,
                                annotation_colors = annotation_map,
                                hm_title = get_hm_title(
                                  normalization_type = "GR",
                                  metric = "hsa_score",
                                  dataset_name = "Co-dilution data"))
hm_3 <- output[["heatmap"]]
patchwork::wrap_elements(hm_3[["gtable"]])

Plot pretty heatmap with annotations for combo data

Description

Plot pretty heatmap with annotations for combo data

Usage

pheatmap_with_anno_combo(
  dt_scores,
  normalization_type = "GR",
  metric = "hsa_score",
  fit_source = "gDR",
  hm_title = NA,
  colors_vec = NULL,
  no_breaks = 50,
  cluster_rows = TRUE,
  cluster_cols = TRUE,
  distfun = compute_distances,
  annotation_row = NULL,
  annotation_col = NULL,
  annotation_colors = NULL,
  max_hm_lbl_length = .get_setting("MAX_HM_LBL_LENGTH")
)

Arguments

dt_scores

data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt and combo SummarizedExperiment
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "hsa_score"("Bliss Excess GR" or "Bliss Excess RV" - respectively depending on normalization_type), "bliss_score" ("Bliss Score GR" or "Bliss Score RV")
fit_source

string source name for metrics
hm_title

string plot title
colors_vec

character vector of colors (valid name or hex) used in heatmap; note that the first color will be assigned to the min value, and the last one - to the max
no_breaks

numeric number of breaks on scale used for mapping values to colors
cluster_rows

logical flag whether rows should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
cluster_cols

logical flag whether columns should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
distfun

function used to compute the distance (dissimilarity) between both rows and columns; used for the dendrogram when cluster_rows or cluster_cols is set to TRUE the default is compute_distances using Spearman method.
annotation_row

data.table that specifies the annotations shown on left side of the heatmap. Each row defines the features for a specific row. The rows in the data and in the annotation are matched using corresponding combination of names from DrugName and DrugName_2 columns. Both columns are required. Note that color schemes takes into account if variable is continuous or discrete.
annotation_col

data.table that specifies the annotations shown above the heatmap. Each row defines the features for a specific column. The columns in the data and in the annotation are matched using corresponding names from the required CellLineName column. Note that color schemes takes into account if variable is continuous or discrete.
annotation_colors

named list for specifying annotation_col and annotation_row track colors manually; note list is named with annotation name (column names of annotation_row - without DrugName and DrugName_2, and column names of annotation_col - without CellLineName), each list item is named vector with a valid color name for each value described in annotation_row) and in annotation_col) - respectively. Not described elements will be colored in default.
max_hm_lbl_length

numeric value for the maximum number of characters in the label; if set to Inf, no trimming will be performed; for better readability, it is recommended to use the default number.

Value

A named list with elements:

  • data a list containing the information visualized in the heatmap:

    • matrix data shown in the heatmap for the selected metric.

    • annotation_row a table with row annotations (for DrugName), if provided.

    • annotation_col a table with column annotations (for CellLineName), if provided.

  • pheatmap object containing the heatmap itself.

Author(s)

Janina Smoła [email protected]

See Also

pheatmap

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_scores <- gDRutils::convert_se_assay_to_dt(se = se,
                                              assay_name = "scores")
annotation_manual_col <-
  unique(dt_scores[, c("CellLineName", "Tissue"), with = FALSE])
annotation_manual_row <-
  unique(dt_scores[, c("DrugName", "DrugName_2", "drug_moa", "drug_moa_2"), with = FALSE])
annotation_map <-
  get_ann_color_map(unique(dt_scores[, c("Tissue", "drug_moa", "drug_moa_2"), with = FALSE]))

output <- pheatmap_with_anno_combo(dt_scores = dt_scores,
                                   normalization_type = "RV",
                                   metric = "bliss_score",
                                   colors_vec = c("darkblue", "grey90", "darkred"),
                                   annotation_row = annotation_manual_row,
                                   annotation_col = annotation_manual_col,
                                   annotation_colors = annotation_map)
hm_1 <- output[["heatmap"]]
patchwork::wrap_elements(hm_1[["gtable"]])

annotation_manual <- data.table::data.table(
  CellLineName =
    c("cellline_AA", "cellline_EA", "cellline_IB", "cellline_MC", "cellline_BC"),
  mut_A = c(1, 1, 1, 0, 0),
  mut_B = c("yes", "yes", "no", "no", "no")
)

annotation_map <- list(
  mut_A = c("1" = "coral", "0" = "cadetblue"),
  mut_B = c("yes" = "black", "no" = "grey90")
)

output <- pheatmap_with_anno_combo(dt_scores = dt_scores,
                                   cluster_cols = FALSE,
                                   annotation_col = annotation_manual,
                                   annotation_colors = annotation_map,
                                   hm_title = get_hm_title(
                                     normalization_type = "GR",
                                     metric = "hsa_score",
                                     dataset_name = "Combo Matrix - combo data"))
hm_2 <- output[["heatmap"]]
patchwork::wrap_elements(hm_2[["gtable"]])

Plot heatmap with annotations for combo metrics data

Description

Plots a heatmap for metrics data (e.g., "xc50", "x_inf") covering both Single Agents and Combinations. Rows are sorted alphabetically by the primary drug, prioritizing Single Agent conditions first, followed by combinations sorted by co-treatment name and concentration.

Usage

pheatmap_with_anno_combo_metrics(
  dt_metrics,
  dt_metrics_capped = NULL,
  normalization_type = "GR",
  metric = "x_mean",
  fit_source = "gDR",
  hm_title = NA,
  colors_vec = NULL,
  no_breaks = 50,
  cluster_rows = FALSE,
  cluster_cols = TRUE,
  distfun = compute_distances,
  annotation_col = NULL,
  annotation_colors = NULL,
  max_hm_lbl_length = .get_setting("MAX_HM_LBL_LENGTH")
)

Arguments

dt_metrics

data.table representing data from the metrics assay, outputted by gDRutils::convert_se_assay_to_dt. Must contain DrugName, DrugName_2, cotrt_value, and CellLineName.
dt_metrics_capped

data.table representing data from the "Metrics" assay, the same as dt_metrics but with capped values with gDRutils::cap_assay_infinities
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean"); but the values from any numeric column can be displayed.
fit_source

string source name for metrics
hm_title

string plot title
colors_vec

character vector of colors (valid name or hex) used in heatmap; note that the first color will be assigned to the min value, and the last one - to the max
no_breaks

numeric number of breaks on scale used for mapping values to colors
cluster_rows

logical flag whether rows should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
cluster_cols

logical flag whether columns should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
distfun

function used to compute the distance (dissimilarity) between both rows and columns; used for the dendrogram when cluster_rows or cluster_cols is set to TRUE the default is compute_distances using Spearman method.
annotation_col

data.table that specifies the annotations shown above the heatmap. Each row defines the features for a specific column. The columns in the data and in the annotation are matched using corresponding names from the required CellLineName column. Note that color schemes takes into account if variable is continuous or discrete.
annotation_colors

named list for specifying annotation_col and annotation_row track colors manually; note list is named with annotation name (column names of annotation_row - without DrugName and column names of annotation_col - without CellLineName), each list item is named vector with valid color name for each value described in annotation_row and in annotation_col - respectively. Not described elements will be colored in default.
max_hm_lbl_length

numeric value for the maximum number of characters in the label; if set to Inf, no trimming will be performed; for better readability, it is recommended to use the default number.

Value

A named list with elements:

  • data a list containing the information visualized in the heatmap:

    • matrix data shown in the heatmap for the selected metric.

    • annotation_row a table with row annotations (auto-generated).

    • annotation_col a table with column annotations, if provided.

  • heatmap the pheatmap object.

Author(s)

Bartosz Czech [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix_small")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Metrics")

output <- pheatmap_with_anno_combo_metrics(dt_metrics = dt_metrics,
                                           normalization_type = "RV",
                                           metric = "x_mean")
hm_1 <- output[["heatmap"]]
patchwork::wrap_elements(hm_1[["gtable"]])

annotation_manual <- data.table::data.table(
  CellLineName =
    c("cellline_AA", "cellline_GB", "cellline_HB"),
  mut_A = c(0, 1, 1),
  mut_B = c("yes", "yes", "no")
)
annotation_map <- list(
  mut_A = c("1" = "coral", "0" = "cadetblue"),
  mut_B = c("yes" = "black", "no" = "grey90")
)

output <- pheatmap_with_anno_combo_metrics(dt_metrics = dt_metrics,
                                           metric = "ec50",
                                           normalization_type = "GR",
                                           annotation_col = annotation_manual,
                                           annotation_colors = annotation_map,
                                           hm_title = get_hm_title(
                                             normalization_type = "GR",
                                             metric = "EC50 for combo data ",
                                             dataset_name = "Combo data"))



mae <- gDRutils::get_synthetic_data("combo_triple")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Metrics",
                                               merge_additional_variables = TRUE)

output <- pheatmap_with_anno_combo_metrics(dt_metrics = dt_metrics,
                                           normalization_type = "RV",
                                           metric = "x_mean")
hm_4 <- output[["heatmap"]]
patchwork::wrap_elements(hm_4[["gtable"]])

Plot pretty heatmap with annotations for single-agent data

Description

Plot pretty heatmap with annotations for single-agent data

Usage

pheatmap_with_anno_sa(
  dt_metrics,
  dt_metrics_capped = NULL,
  normalization_type = "GR",
  metric = "xc50",
  fit_source = "gDR",
  hm_title = NA,
  colors_vec = NULL,
  no_breaks = 50,
  cluster_rows = TRUE,
  cluster_cols = TRUE,
  distfun = compute_distances,
  annotation_row = NULL,
  annotation_col = NULL,
  annotation_colors = NULL,
  max_hm_lbl_length = .get_setting("MAX_HM_LBL_LENGTH")
)

Arguments

dt_metrics

data.table representing data from the "Metrics" assay, outputted by gDRutils::convert_se_assay_to_dt and single-agent SummarizedExperiment
dt_metrics_capped

data.table representing data from the "Metrics" assay, the same as dt_metrics but with capped values with gDRutils::cap_assay_infinities
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean"); but the values from any numeric column can be displayed.
fit_source

string source name for metrics
hm_title

string plot title
colors_vec

character vector of colors (valid name or hex) used in heatmap; note that the first color will be assigned to the min value, and the last one - to the max
no_breaks

numeric number of breaks on scale used for mapping values to colors
cluster_rows

logical flag whether rows should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
cluster_cols

logical flag whether columns should be clustered; the dendrogram will not be shown for the matrix with any dimension greater than 200.
distfun

function used to compute the distance (dissimilarity) between both rows and columns; used for the dendrogram when cluster_rows or cluster_cols is set to TRUE the default is compute_distances using Spearman method.
annotation_row

data.table that specifies the annotations shown on left side of the heatmap. Each row defines the features for a specific row. The rows in the data and in the annotation are matched using corresponding names from the required DrugName column. Note that color schemes takes into account if variable is continuous or discrete.
annotation_col

data.table that specifies the annotations shown above the heatmap. Each row defines the features for a specific column. The columns in the data and in the annotation are matched using corresponding names from the required CellLineName column. Note that color schemes takes into account if variable is continuous or discrete.
annotation_colors

named list for specifying annotation_col and annotation_row track colors manually; note list is named with annotation name (column names of annotation_row - without DrugName and column names of annotation_col - without CellLineName), each list item is named vector with valid color name for each value described in annotation_row and in annotation_col - respectively. Not described elements will be colored in default.
max_hm_lbl_length

numeric value for the maximum number of characters in the label; if set to Inf, no trimming will be performed; for better readability, it is recommended to use the default number.

Value

A named list with elements:

  • data a list containing the information visualized in the heatmap:

    • matrix data shown in the heatmap for the selected metric.

    • annotation_row a table with row annotations (for DrugName), if provided.

    • annotation_col a table with column annotations (for CellLineName), if provided.

  • pheatmap object containing the heatmap itself.

Author(s)

Janina Smoła [email protected]

See Also

pheatmap::pheatmap

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Metrics")
dt_averaged <- gDRutils::convert_se_assay_to_dt(se = se,
                                                assay_name = "Averaged")
dt_metrics_capped <-
  gDRutils::cap_assay_infinities(
    conc_assay_dt = dt_averaged,
    assay_dt = dt_metrics,
    experiment_name = gDRutils::get_supported_experiments("sa"),
    col = "xc50",
    capping_fold = 5)

output <- pheatmap_with_anno_sa(dt_metrics = dt_metrics)
hm_0 <- output[["heatmap"]]
patchwork::wrap_elements(hm_0[["gtable"]])

output <- pheatmap_with_anno_sa(dt_metrics = dt_metrics,
                                dt_metrics_capped = dt_metrics_capped)
hm_1 <- output[["heatmap"]]
patchwork::wrap_elements(hm_1[["gtable"]])

annotation_manual_col <-
  unique(dt_metrics[, c("CellLineName", "Tissue"), with = FALSE])
annotation_manual_row <-
  unique(dt_metrics[, c("DrugName", "drug_moa"), with = FALSE])
annotation_map <-
  get_ann_color_map(unique(dt_metrics[, c("Tissue", "drug_moa"), with = FALSE]))

output <- pheatmap_with_anno_sa(dt_metrics = dt_metrics,
                                normalization_type = "RV",
                                metric = "x_mean",
                                colors_vec = c("darkblue", "grey90"),
                                annotation_row = annotation_manual_row,
                                annotation_col = annotation_manual_col,
                                annotation_colors = annotation_map)
hm_2 <- output[["heatmap"]]
patchwork::wrap_elements(hm_2[["gtable"]])

annotation_manual <- data.table::data.table(
  CellLineName =
    c("cellline_AA", "cellline_EA", "cellline_IB", "cellline_MC", "cellline_BC"),
  mut_A = c(1, 1, 1, 0, 0),
  mut_B = c("yes", "yes", "no", "no", "no")
)
annotation_map <- list(
  mut_A = c("1" = "coral", "0" = "cadetblue"),
  mut_B = c("yes" = "black", "no" = "grey90")
)

output <- pheatmap_with_anno_sa(dt_metrics = dt_metrics,
                                annotation_col = annotation_manual,
                                annotation_colors = annotation_map,
                                hm_title = get_hm_title(
                                  normalization_type = "GR",
                                  metric = "hsa_score",
                                  dataset_name = "Combo Matrix - combo data"))
hm_3 <- output[["heatmap"]]
patchwork::wrap_elements(hm_3[["gtable"]])

Plot boxplot for metric values grouped by metadata from DepMap

Description

Plot boxplot for metric values grouped by metadata from DepMap

Usage

plot_boxplot_meta(
  dt_response,
  dt_depmap,
  selected_feat_meta_col,
  with_1_item_grp = TRUE,
  max_x_lbl_length = 60,
  with_inf = FALSE
)

Arguments

dt_response

data.table with experimental response data (rows are samples) for one metric outputted by one of functions: prep_dt_response_metric_sa, prep_dt_response_dose_sa, prep_dt_response_scores or prep_dt_response_metric_diff,
dt_depmap

data.table with dependent variables data load from DepMap - for one metadata; (rows are samples, columns are metadata levels); outputted by prep_dt_depmap_meta
selected_feat_meta_col

string with the name of the selected metadata from dt_depmap (will be used as a plot title)
with_1_item_grp

logical flag indicating whether to show group with only one item
max_x_lbl_length

numeric value for the maximum number of characters in the x-axis label
with_inf

a logical flag indicating whether infinite values should be shown on boxplots

Value

ggplot object containing boxplots for variable levels

Author(s)

Janina Smoła [email protected]

Examples

mae_prism <- gDRutils::get_synthetic_data("prism")
se_prism <- mae_prism[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_prism, "Metrics")
dt_response <- prep_dt_response_metric_sa(
  dt_metrics = dt_metrics,
  d_name = unique(dt_metrics[["DrugName"]])[1],
  normalization_type = "RV", metric = "x_mean")
meta_data_path <- system.file("depmap_data/Model.csv.gz", package = "gDRtestData")
obj_depmap_meta <- prep_dt_depmap_meta(meta_data_path, metadata_col = "OncotreeLineage")
plot_boxplot_meta(
  dt_response = dt_response,
  dt_depmap = obj_depmap_meta[["dt_depmap"]],
  selected_feat_meta_col = obj_depmap_meta[["selected_feat_meta_col"]])

Plot box plots for metric for combo data grouped by selected variable

Description

Plot box plots for metric for combo data grouped by selected variable

Usage

plot_boxplot_metric_combo(
  dt_scores,
  group_var,
  normalization_type = "GR",
  metric = "hsa_score",
  fit_source = "gDR",
  grouped_flag = FALSE,
  colored_pts_flag = FALSE,
  colors_vec = NULL
)

Arguments

dt_scores

data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt(se, "scores") and combo SummarizedExperiment
group_var

string name of group variable; one of: "CellLineName" or "DrugName"; for group_var set as "DrugName" points will be grouped by drug combinations "DrugName" x "DrugName_2"
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "hsa_score"("Bliss Excess GR" or "Bliss Excess RV" - respectively depending on normalization_type) or "bliss_score" ("Bliss Score GR" or "Bliss Score RV")
fit_source

string source name for metrics
grouped_flag

a logical flag whether the boxplots should be grouped and colored by Tissue for group_var set as "CellLineName" and drug_moa - for "DrugName"
colored_pts_flag

a logical flag whether the points should be colored by grouped variable - for group_var equal "CellLineName" points will be colored by "DrugName" and similarly vice versa
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used

Value

ggplot object containing boxplots for selected combo metric grouped by group_var

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]

dt_scores <- gDRutils::convert_se_assay_to_dt(se = se,
                                              assay_name = "scores")

plot_boxplot_metric_combo(dt_scores,
                          group_var = "DrugName")

plot_boxplot_metric_combo(dt_scores,
                          group_var = "DrugName",
                          colored_pts_flag = TRUE,
                          colors_vec = "grey")

plot_boxplot_metric_combo(dt_scores,
                          group_var = "CellLineName",
                          metric = "hsa_score",
                          normalization_type = "RV",
                          grouped_flag = TRUE)

plot_boxplot_metric_combo(
  dt_scores,
  group_var = "CellLineName",
  metric = "hsa_score",
  grouped_flag = TRUE,
  colors_vec = c("deeppink", "darkblue"))

Plot box plots for metric for combo data grouped by cell line names

Description

Plot box plots for metric for combo data grouped by cell line names

Usage

plot_boxplot_metric_combo_by_CLs(
  dt_scores,
  normalization_type = "GR",
  metric = "hsa_score",
  fit_source = "gDR",
  grouped_flag = FALSE,
  colored_pts_flag = FALSE,
  colors_vec = NULL
)

Arguments

dt_scores

data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt(se, "scores") and combo SummarizedExperiment
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "hsa_score"("Bliss Excess GR" or "Bliss Excess RV" - respectively depending on normalization_type) or "bliss_score" ("Bliss Score GR" or "Bliss Score RV")
fit_source

string source name for metrics
grouped_flag

a logical flag whether the boxplots should be grouped and colored by Tissue for group_var set as "CellLineName" and drug_moa - for "DrugName"
colored_pts_flag

a logical flag whether the points should be colored by grouped variable - for group_var equal "CellLineName" points will be colored by "DrugName" and similarly vice versa
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used

Value

ggplot object containing boxplots for selected combo metric grouped by cell line names

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]

dt_scores <- gDRutils::convert_se_assay_to_dt(se = se,
                                              assay_name = "scores")

plot_boxplot_metric_combo_by_CLs(dt_scores)

plot_boxplot_metric_combo_by_CLs(dt_scores,
                                 normalization_type = "RV",
                                 grouped_flag = TRUE)

plot_boxplot_metric_combo_by_CLs(dt_scores,
                                 metric = "bliss_score",
                                 colors_vec = "gold")

plot_boxplot_metric_combo_by_CLs(
  dt_scores,
  metric = "bliss_score",
  colored_pts_flag = TRUE)

plot_boxplot_metric_combo_by_CLs(
  dt_scores,
  metric = "hsa_score",
  normalization_type = "RV",
  grouped_flag = TRUE,
  colors_vec = c("tomato", "darkgreen", "orange", "darkblue"))

Plot box plots for metric for combo data grouped by drug names

Description

Plot box plots for metric for combo data grouped by drug names

Usage

plot_boxplot_metric_combo_by_drugs(
  dt_scores,
  normalization_type = "GR",
  metric = "hsa_score",
  fit_source = "gDR",
  colored_pts_flag = FALSE,
  colors_vec = NULL
)

Arguments

dt_scores

data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt(se, "scores") and combo SummarizedExperiment
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "hsa_score"("Bliss Excess GR" or "Bliss Excess RV" - respectively depending on normalization_type) or "bliss_score" ("Bliss Score GR" or "Bliss Score RV")
fit_source

string source name for metrics
colored_pts_flag

a logical flag whether the points should be colored by grouped variable - for group_var equal "CellLineName" points will be colored by "DrugName" and similarly vice versa
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used

Value

ggplot object containing boxplots for selected combo metric grouped by drug names

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]

dt_scores <- gDRutils::convert_se_assay_to_dt(se = se,
                                              assay_name = "scores")

plot_boxplot_metric_combo_by_drugs(dt_scores)

plot_boxplot_metric_combo_by_drugs(dt_scores,
                                   normalization_type = "RV",
                                   colors_vec = "gold")

plot_boxplot_metric_combo_by_drugs(dt_scores,
                                   metric = "bliss_score",
                                   normalization_type = "RV",
                                   colored_pts_flag = TRUE)

Plot box plots for metric for combo data grouped by selected variable

Description

Plot box plots for metric for combo data grouped by selected variable

Usage

plot_boxplot_metric_combo_by_grp(
  dt_scores,
  selection_var,
  selection_name,
  group_var,
  group_names = NULL,
  normalization_type = "GR",
  metric = "hsa_score",
  fit_source = "gDR",
  named_n = 5,
  named_n_mode = "top",
  grouped_flag = FALSE,
  colors_vec = NULL
)

Arguments

dt_scores

data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt(se, "scores") and combo SummarizedExperiment
selection_var

string name of selected main variable - one value from column "CellLineName" or "DrugName"
selection_name

string name of selected variable value from column selection_var to filter data for plotting
group_var

string name of group variable; should to be not numeric variable from dt_metrics different than selection_var and not containing unique values for each row
group_names

character vector with names to subset from column group_var; if NULL then all values will be plotted
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "hsa_score"("Bliss Excess GR" or "Bliss Excess RV" - respectively depending on normalization_type) or "bliss_score" ("Bliss Score GR" or "Bliss Score RV")
fit_source

string source name for metrics
named_n

number of points to label based on the highest or lowest metric values; if group_var is "DrugName", points are labeled by "CellLineName" and similarly vice versa
named_n_mode

string determines whether the labels are applied to the highest or lowest values of metric; one of: "top" or "bottom"
grouped_flag

logical flag whether the boxplots should be colored by group_var
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used

Value

ggplot object containing boxplots for selected combo metric grouped by selected variable

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]

dt_scores <- gDRutils::convert_se_assay_to_dt(se = se,
                                              assay_name = "scores")
invisible(dt_scores[, Tissue_grp := data.table::fifelse(Tissue == "tissue_w",
                                                        "tissue_w",
                                                        "tissue_other")])

plot_boxplot_metric_combo_by_grp(dt_scores,
                                 selection_var = "DrugName",
                                 selection_name = c("drug_001", "drug_021"),
                                 group_var = "Tissue")

plot_boxplot_metric_combo_by_grp(dt_scores,
                                 selection_var = "DrugName",
                                 selection_name = c("drug_001", "drug_021"),
                                 group_var = "Tissue_grp",
                                 grouped_flag = TRUE,
                                 colors_vec = c("darkblue", "deeppink"))

plot_boxplot_metric_combo_by_grp(dt_scores,
                                 selection_var = "DrugName",
                                 selection_name = c("drug_001", "drug_021"),
                                 group_var = "Tissue",
                                 grouped_flag = TRUE)

plot_boxplot_metric_combo_by_grp(dt_scores,
                                 selection_var = "DrugName",
                                 selection_name = c("drug_001", "drug_021"),
                                 group_var = "Tissue",
                                 colors_vec = c("darkblue", "orange"))

Plot box plots for metric for single-agent data grouped by selected variable

Description

Plot box plots for metric for single-agent data grouped by selected variable

Usage

plot_boxplot_metric_sa(
  dt_metrics,
  group_var,
  normalization_type = "GR",
  metric = "xc50",
  fit_source = "gDR",
  grouped_flag = FALSE,
  colored_pts_flag = FALSE,
  colors_vec = NULL,
  with_inf = FALSE
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
group_var

string name of group variable; one of: "CellLineName" or "DrugName"
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean"), but the values from any numeric column can be displayed
fit_source

string source name for metrics
grouped_flag

a logical flag whether the boxplots should be grouped and colored by Tissue for group_var set as "CellLineName" and drug_moa - for "DrugName"
colored_pts_flag

a logical flag whether the points should be colored by grouped variable - for group_var equal "CellLineName" points will be colored by "DrugName" and similarly vice versa
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used
with_inf

a logical flag indicating whether infinite values should be shown on boxplots

Value

ggplot object containing boxplots for selected single-agent metric grouped by group_var

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")

plot_boxplot_metric_sa(dt_metrics,
                       group_var = "CellLineName")

plot_boxplot_metric_sa(dt_metrics,
                       group_var = "DrugName")

plot_boxplot_metric_sa(dt_metrics,
                       group_var = "DrugName",
                       metric = "x_AOC_range",
                       colors_vec = "grey",
                       colored_pts_flag = TRUE)

plot_boxplot_metric_sa(dt_metrics,
                       group_var = "CellLineName",
                       normalization_type = "RV",
                       metric = "x_max",
                       colors_vec = c("gold", "darkorange", "darkcyan", "darkblue"),
                       grouped_flag = TRUE)

Plot box plots for metric for single-agent data grouped by cell line names

Description

Plot box plots for metric for single-agent data grouped by cell line names

Usage

plot_boxplot_metric_sa_by_CLs(
  dt_metrics,
  normalization_type = "GR",
  metric = "xc50",
  fit_source = "gDR",
  grouped_flag = FALSE,
  colored_pts_flag = FALSE,
  colors_vec = NULL,
  with_inf = FALSE
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean"), but the values from any numeric column can be displayed
fit_source

string source name for metrics
grouped_flag

a logical flag whether the boxplots should be grouped and colored by Tissue for group_var set as "CellLineName" and drug_moa - for "DrugName"
colored_pts_flag

a logical flag whether the points should be colored by grouped variable - for group_var equal "CellLineName" points will be colored by "DrugName" and similarly vice versa
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used
with_inf

a logical flag indicating whether infinite values should be shown on boxplots

Value

ggplot object containing boxplots for selected single-agent metric grouped by cellline names

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")

plot_boxplot_metric_sa_by_CLs(dt_metrics)

plot_boxplot_metric_sa_by_CLs(dt_metrics,
                              grouped_flag = TRUE)

plot_boxplot_metric_sa_by_CLs(dt_metrics,
                              with_inf = TRUE)

plot_boxplot_metric_sa_by_CLs(dt_metrics,
                              metric = "x_AOC",
                              colors_vec = "gold")

plot_boxplot_metric_sa_by_CLs(
  dt_metrics,
  metric = "x_max",
  colored_pts_flag = TRUE)

plot_boxplot_metric_sa_by_CLs(
  dt_metrics,
  metric = "x_mean",
  grouped_flag = TRUE,
  colors_vec = c("deeppink", "darkcyan", "orange", "darkblue"))

Plot box plots for metric for single-agent data grouped by drug names

Description

Plot box plots for metric for single-agent data grouped by drug names

Usage

plot_boxplot_metric_sa_by_drugs(
  dt_metrics,
  normalization_type = "GR",
  metric = "xc50",
  fit_source = "gDR",
  grouped_flag = FALSE,
  colored_pts_flag = FALSE,
  colors_vec = NULL,
  with_inf = FALSE
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean"), but the values from any numeric column can be displayed
fit_source

string source name for metrics
grouped_flag

a logical flag whether the boxplots should be grouped and colored by Tissue for group_var set as "CellLineName" and drug_moa - for "DrugName"
colored_pts_flag

a logical flag whether the points should be colored by grouped variable - for group_var equal "CellLineName" points will be colored by "DrugName" and similarly vice versa
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used
with_inf

a logical flag indicating whether infinite values should be shown on boxplots

Value

ggplot object containing boxplots for selected single-agent metric grouped by drug names

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")

plot_boxplot_metric_sa_by_drugs(dt_metrics,
                                normalization_type = "RV")

plot_boxplot_metric_sa_by_drugs(dt_metrics,
                                normalization_type = "RV",
                                grouped_flag = TRUE)

plot_boxplot_metric_sa_by_drugs(dt_metrics,
                                metric = "x_AOC",
                                colors_vec = "gold")

plot_boxplot_metric_sa_by_drugs(
  dt_metrics,
  metric = "x_max",
  colored_pts_flag = TRUE)

plot_boxplot_metric_sa_by_drugs(
  dt_metrics,
  metric = "x_mean",
  grouped_flag = TRUE,
  colors_vec = c("deeppink", "darkcyan", "orange", "darkblue", "limegreen"))

Plot box plots for metric for single-agent data grouped by selected variable

Description

Plot box plots for metric for single-agent data grouped by selected variable

Usage

plot_boxplot_metric_sa_by_grp(
  dt_metrics,
  selection_var,
  selection_name,
  group_var,
  group_names = NULL,
  normalization_type = "GR",
  metric = "xc50",
  fit_source = "gDR",
  named_n = 5,
  named_n_mode = "bottom",
  grouped_flag = FALSE,
  colors_vec = NULL,
  with_inf = FALSE
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
selection_var

string name of selected main variable - one value from column "CellLineName" or "DrugName"
selection_name

string name of selected variable value from column selection_var to filter data for plotting
group_var

string name of group variable; should to be not numeric variable from dt_metrics different than selection_var and not containing unique values for each row
group_names

character vector with names to subset from column group_var; if NULL then all values will be plotted
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean"), but the values from any numeric column can be displayed
fit_source

string source name for metrics
named_n

number of points to label based on the highest or lowest metric values; if group_var is "DrugName", points are labeled by "CellLineName" and similarly vice versa
named_n_mode

string determines whether the labels are applied to the highest or lowest values of metric; one of: "top" or "bottom"
grouped_flag

logical flag whether the boxplots should be colored by group_var
colors_vec

character vector with colors (name or hex value) to color boxplots; for grouped_flag set as FALSE only first from vector will be used
with_inf

a logical flag indicating whether infinite values should be shown on boxplots

Value

ggplot object containing boxplots for selected single-agent metric grouped by selected variable

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
invisible(dt_metrics[, Tissue_grp := data.table::fifelse(Tissue == "tissue_w",
                                                         "tissue_w",
                                                         "tissue_other")])

plot_boxplot_metric_sa_by_grp(dt_metrics,
                              selection_var = "DrugName",
                              selection_name = "drug_001",
                              group_var = "Tissue")

plot_boxplot_metric_sa_by_grp(dt_metrics,
                              selection_var = "DrugName",
                              selection_name = "drug_001",
                              group_var = "Tissue_grp")

plot_boxplot_metric_sa_by_grp(dt_metrics,
                              selection_var = "DrugName",
                              selection_name = "drug_001",
                              group_var = "Tissue",
                              with_inf = TRUE)

plot_boxplot_metric_sa_by_grp(dt_metrics,
                              selection_var = "DrugName",
                              selection_name = "drug_001",
                              group_var = "Tissue",
                              named_n = 0)

plot_boxplot_metric_sa_by_grp(dt_metrics,
                              selection_var = "DrugName",
                              selection_name = "drug_001",
                              group_var = "Tissue",
                              grouped_flag = TRUE)

plot_boxplot_metric_sa_by_grp(dt_metrics,
                              selection_var = "DrugName",
                              selection_name = "drug_001",
                              group_var = "Tissue",
                              colors_vec = c("darkblue", "deeppink"))

Plot boxplot for categorical features

Description

Plot boxplot for categorical features

Usage

plot_boxplot_num(
  dt_response,
  dt_depmap,
  selected_feat,
  selected_feat_meta_col = NULL
)

Arguments

dt_response

data.table with experimental response data (rows are samples) for one metric outputted by one of functions: prep_dt_response_metric_sa, prep_dt_response_dose_sa, prep_dt_response_scores or prep_dt_response_metric_diff,
dt_depmap

data.table with dependent variables data loaded from DepMap - for one feature; (rows are samples, columns are features). outputted by prep_dt_depmap_feat
selected_feat

string with name of selected feature from dt_depmap
selected_feat_meta_col

string with the name of a feature column in DepMap (will be used as a plot title) that has 0-1 values only (categorical character but without relation one-to-one for ids and feature)

Value

ggplot object containing boxplots for variable levels

Author(s)

Janina Smoła [email protected]

Examples

mae_prism <- gDRutils::get_synthetic_data("prism")
se_prism <- mae_prism[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_prism, "Metrics")
dt_response <- prep_dt_response_metric_sa(
  dt_metrics = dt_metrics,
  d_name = unique(dt_metrics[["DrugName"]])[1],
  normalization_type = "RV", metric = "x_mean")
meta_data_path <- system.file("depmap_data/Model.csv.gz", package = "gDRtestData")
obj_depmap_meta <- prep_dt_depmap_meta(meta_data_path, metadata_col = "OncotreeLineage")
plot_boxplot_num(
  dt_response = dt_response,
  dt_depmap = obj_depmap_meta[["dt_depmap"]],
  selected_feat = names(obj_depmap_meta[["dt_depmap"]])[3])

Plot panel with boxplots

Description

Plot panel with boxplots

Usage

plot_boxplot_num_panel(
  dt_response,
  dt_depmap,
  selected_feats,
  selected_feat_meta_col = NULL,
  ncol = NULL
)

Arguments

dt_response

data.table with experimental response data (rows are samples) for one metric outputted by one of functions: prep_dt_response_metric_sa, prep_dt_response_dose_sa, prep_dt_response_scores or prep_dt_response_metric_diff,
dt_depmap

data.table with dependent variables data loaded from DepMap - for one feature; (rows are samples, columns are features). outputted by prep_dt_depmap_feat
selected_feats

character vector with names of selected features from dt_depmap
selected_feat_meta_col

string with the name of a feature column in DepMap (will be used as a plot title) that has 0-1 values only (categorical character but without relation one-to-one for ids and feature)
ncol

number of plot column in panel

Value

ggplot object containing boxplots for variable levels (0-1)

Author(s)

Janina Smoła [email protected]

Examples

mae_prism <- gDRutils::get_synthetic_data("prism")
se_prism <- mae_prism[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_prism, "Metrics")
dt_response <- prep_dt_response_metric_sa(
  dt_metrics = dt_metrics,
  d_name = unique(dt_metrics[["DrugName"]])[1],
  normalization_type = "RV", metric = "x_mean")
meta_data_path <- system.file("depmap_data/Model.csv.gz", package = "gDRtestData")
obj_depmap_meta <- prep_dt_depmap_meta(meta_data_path, metadata_col = "OncotreeLineage")
plot_boxplot_num_panel(
  dt_response = dt_response,
  dt_depmap = obj_depmap_meta[["dt_depmap"]],
  selected_feats = names(obj_depmap_meta[["dt_depmap"]])[3:5])

Plot Chemical Genomics Screen GSEA Results

Description

Generates a ggplot2 visualization of chemical genomics screening data, highlighting GSEA results.

Usage

plot_cgs_ranking(
  results,
  cl_name,
  metric,
  padj_threshold = 0.1,
  top_results_no_sig = 5,
  max_results_with_sig = 15
)

Arguments

results

A list object returned from 'analyze_cgs'.
cl_name

A string specifying the cell line included in the results to prepare a visualization.
metric

A string specifying the metric included in the results to prepare a visualization.
padj_threshold

A numeric value specifying the threshold for filtering significant GSEA results based on adjusted p-value.
top_results_no_sig

A numeric value specifying the number of top results to plot if there are no significant values.
max_results_with_sig

A numeric value specifying the maximum number of results to plot when there are more than this number of significant values.

Value

A ggplot2 object with cgs results

Author(s)

Bartosz Czech [email protected]

Examples

dt_metrics <- qs2::qs_read(system.file("testdata/cgs_data.qs2", package = "gDRplots"))
results <- analyze_cgs(dt_metrics, metrics = c("xc50"), cl_name = "CellLineName_1")
plot_cgs_ranking(results,
  cl_name = "CellLineName_1",
  metric = "xc50",
  padj_threshold = 0.1,
  top_results_no_sig = 5,
  max_results_with_sig = 15)

Plot line plot of combination index

Description

Plot line plot of combination index

Usage

plot_combination_index(
  dt_excess,
  dt_isobolograms,
  drug1_name,
  drug2_name,
  cl_name,
  normalization_type = "GR",
  iso_levels = c("0.25", "0.5", "0.75"),
  colors_vec_iso = NULL
)

Arguments

dt_excess

data.table representing data from the excess assay, outputted by gDRutils::convert_se_assay_to_dt(se, "excess") and combo SummarizedExperiment
dt_isobolograms

data.table representing data from the isobolograms assay, outputted by gDRutils::convert_se_assay_to_dt(se, "isobolograms") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_name

string with cell line to be plotted (identifiers CellLineName)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
iso_levels

character vector with isobologram levels to be selected
colors_vec_iso

character vector of colors (valid name or hex) used for the isolines; the default is the dark red-orange palette

Value

ggplot object containing combination index plot at different ratios of the two drugs

Examples

cl_name <- "cellline_BC"
drug1_name <- "drug_001"
drug2_name <- "drug_026"

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_excess <- gDRutils::convert_se_assay_to_dt(se, "excess")
dt_isobolograms <- gDRutils::convert_se_assay_to_dt(se, "isobolograms")

plot_combination_index(dt_excess,
                       dt_isobolograms,
                       drug1_name, drug2_name,
                       cl_name,
                       normalization_type = "GR")

plot_combination_index(dt_excess,
                       dt_isobolograms,
                       drug1_name, drug2_name,
                       cl_name,
                       normalization_type = "RV",
                       colors_vec_iso = c("darkblue", "darkcyan"))

cl_name <- "cellline_JE"
drug1_name <- "drug_011"
drug2_name <- "drug_026"

plot_combination_index(dt_excess,
                       dt_isobolograms,
                       drug1_name, drug2_name,
                       cl_name,
                       normalization_type = "RV",
                       iso_levels = "0.5")

Plot drug response curves for combo data

Description

Plot drug response curves for combo data

Usage

plot_dose_response_combo(
  dt_average,
  drug1_name,
  drug2_name,
  cl_name,
  normalization_type = "GR",
  colors_vec = NULL,
  split_by_conc = FALSE
)

Arguments

dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and combo SummarizedExperiment
drug1_name

string with drug name to be plotted (identifiers DrugName)
drug2_name

string with co-drug name to be plotted (identifiers DrugName_2)
cl_name

string with cell line name to be plotted (identifiers CellLineName)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
colors_vec

character vector with colors for group_names - name or hex value note that the first color will be assigned to the min value of Concentration_2, and the last one - to the max of Concentration_2; the default is the orange-black palette
split_by_conc

split_by_conc logical flag indicating whether curves for Concentration_2 should be plotted on a single plot or separately

Value

ggplot object containing plot with dose-response curves for combo data

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")

cl_name <- "cellline_BC"
drug1_name <- "drug_011"
drug2_name <- "drug_021"

plot_dose_response_combo(dt_average = dt_average,
                         drug1_name = drug1_name,
                         drug2_name = drug2_name,
                         cl_name = cl_name)

plot_dose_response_combo(dt_average = dt_average,
                         drug1_name = drug1_name,
                         drug2_name = drug2_name,
                         cl_name = cl_name,
                         split_by_conc = TRUE)

plot_dose_response_combo(dt_average = dt_average,
                         drug1_name = drug1_name,
                         drug2_name = drug2_name,
                         cl_name = cl_name,
                         normalization_type = "RV",
                         colors_vec = c("lightblue", "darkblue"))

Plot panel with drug response curves for single-agent data to control quality of the data

Description

Plot panel with drug response curves for single-agent data to control quality of the data

Usage

plot_dose_response_combo_panel(
  dt_average,
  cl_name,
  d_names = NULL,
  normalization_type = "GR",
  colors_vec = NULL
)

Arguments

dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and combo SummarizedExperiment
cl_name

string with cell line name to be plotted (identifiers CellLineName)
d_names

character vector with drug names to be plotted (Drug Name); if NULL - all available drugs will be plotted
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
colors_vec

character vector with colors for group_names - name or hex value note that the first color will be assigned to the min value of Concentration_2, and the last one - to the max of Concentration_2; the default is the orange-black palette

Value

ggplot object containing panel with plot with dose-response curves for selected cell line by drugs

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")

cl_name <- "cellline_IB"

plot_dose_response_combo_panel(dt_average = dt_average,
                               cl_name = cl_name)

d_names <- c("drug_001", "drug_002")
plot_dose_response_combo_panel(dt_average = dt_average,
                               cl_name = cl_name,
                               d_names = d_names)

Plot drug response curves for single-agent data

Description

Plot drug response curves for single-agent data

Usage

plot_dose_response_sa(
  dt_metrics,
  dt_average,
  selection_name,
  group_var,
  group_names = NULL,
  normalization_type = "GR",
  colors_vec = NULL,
  plot_averaged_flag = TRUE,
  plot_fit_flag = TRUE,
  fit_source = "gDR"
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and single-agent SummarizedExperiment
selection_name

string name of selected main variable - one value from column "CellLineName" or "DrugName"
group_var

string name of group variable; one of: "CellLineName" or "DrugName"
group_names

character vector with names to subset from se (the same dim as group_var); if NULL then all values will be plotted
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
colors_vec

character vector with colors for group_names - name or hex value
plot_averaged_flag

logical flag whether plot points with average values
plot_fit_flag

logical flag whether plot points with fitted values
fit_source

string source name for metrics

Value

ggplot object containing plot of dose-response curves

Note

inspired by the grob_SA function written by Marc Hafner

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]
selected_drug <- "drug_002"
group_var <- "CellLineName"
dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
celline_names <- unique(dt_metrics[[group_var]])[1:3]

plot_dose_response_sa(dt_metrics = dt_metrics,
                      dt_average = dt_average,
                      selection_name = selected_drug,
                      group_var = group_var,
                      group_names = celline_names)

plot_dose_response_sa(dt_metrics = dt_metrics,
                      dt_average = NULL,
                      selection_name = selected_drug,
                      group_var = group_var,
                      group_names = celline_names)

selected_cellline <- "cellline_HB"
group_var <- "DrugName"
dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
group_names <- unique(dt_metrics[[group_var]])[1:3]

plot_dose_response_sa(dt_metrics = dt_metrics,
                      dt_average = dt_average,
                      selection_name = selected_cellline,
                      group_var = group_var,
                      group_names = group_names)

Plot drug response curves for single-agent data for selected call lines and drugs

Description

Plot drug response curves for single-agent data for selected call lines and drugs

Usage

plot_dose_response_sa_by_CLs(
  dt_metrics,
  dt_average,
  cellline_name_vec = NULL,
  drug_name_vec = NULL,
  normalization_type = "GR",
  colors_vec = NULL,
  plot_averaged_flag = TRUE,
  plot_fit_flag = TRUE,
  fit_source = "gDR"
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and single-agent SummarizedExperiment
cellline_name_vec

character vector with cell line names to be plotted (Cell Line Name)
drug_name_vec

character vector with drug names to be plotted (Drug Name)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
colors_vec

character vector with colors for group_names - name or hex value
plot_averaged_flag

logical flag whether plot points with average values
plot_fit_flag

logical flag whether plot points with fitted values
fit_source

string source name for metrics

Value

list of ggplot objects containing plots of dose-response curves

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
cellline_name_vec <- unique(dt_metrics[["CellLineName"]])[2:5]
drug_name_vec <- unique(dt_metrics[["DrugName"]])[5:7]

plot_dose_response_sa_by_CLs(dt_metrics = dt_metrics,
                             dt_average = dt_average,
                             cellline_name_vec = cellline_name_vec,
                             drug_name_vec = drug_name_vec,
                             normalization_type = "RV",
                             colors_vec = c("#00008B", "#FF6347", "#4CBB17"))

Plot drug response curves for single-agent data for selected call lines and drugs

Description

Plot drug response curves for single-agent data for selected call lines and drugs

Usage

plot_dose_response_sa_by_drugs(
  dt_metrics,
  dt_average,
  cellline_name_vec = NULL,
  drug_name_vec = NULL,
  normalization_type = "GR",
  colors_vec = NULL,
  plot_averaged_flag = TRUE,
  plot_fit_flag = TRUE,
  fit_source = "gDR"
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and single-agent SummarizedExperiment
cellline_name_vec

character vector with cell line names to be plotted (Cell Line Name)
drug_name_vec

character vector with drug names to be plotted (Drug Name)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
colors_vec

character vector with colors for group_names - name or hex value
plot_averaged_flag

logical flag whether plot points with average values
plot_fit_flag

logical flag whether plot points with fitted values
fit_source

string source name for metrics

Value

list of ggplot objects containing plots of dose-response curves

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
cellline_name_vec <- unique(dt_metrics[["CellLineName"]])[2:5]
drug_name_vec <- unique(dt_metrics[["DrugName"]])[5:7]

plot_dose_response_sa_by_drugs(dt_metrics = dt_metrics,
                               dt_average = dt_average,
                               cellline_name_vec = cellline_name_vec,
                               drug_name_vec = drug_name_vec,
                               normalization_type = "RV",
                               colors_vec = c("#00008B", "#FF6347", "#4CBB17"))

Plot drug response curves for single-agent data to control quality of the data

Description

Plot drug response curves for single-agent data to control quality of the data

Usage

plot_dose_response_sa_qc(
  dt_metrics,
  dt_average,
  cl_name,
  d_name,
  normalization_type = "GR",
  fit_source = "gDR"
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and single-agent SummarizedExperiment
cl_name

string cell line name to be plotted (Cell Line Name)
d_name

string vector with drug name to be plotted (Drug Name)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
fit_source

string source name for metrics

Value

ggplot object containing plot of dose-response curves (observed and fitted values)

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
cl_name <- dt_metrics[["CellLineName"]][1]
d_name <- dt_metrics[["DrugName"]][1]

plot_dose_response_sa_qc(dt_metrics = dt_metrics,
                         dt_average = dt_average,
                         cl_name = cl_name,
                         d_name = d_name)

Plot panel with drug response curves for single-agent data to control quality of the data

Description

Plot panel with drug response curves for single-agent data to control quality of the data

Usage

plot_dose_response_sa_qc_panel(
  dt_metrics,
  dt_average,
  cl_name,
  d_names = NULL,
  normalization_type = "GR",
  fit_source = "gDR"
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Averaged") and single-agent SummarizedExperiment
cl_name

string cell line name to be plotted (Cell Line Name)
d_names

character vector with drug names to be plotted (Drug Name); if NULL - all available drugs will be plotted
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
fit_source

string source name for metrics

Value

ggplot object with panel with plots of dose-response curves for selected cell line by drugs (observed and fitted values)

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[1]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
cl_name <- dt_metrics[["CellLineName"]][1]
d_names <- unique(dt_metrics[["DrugName"]])[1:5]

plot_dose_response_sa_qc_panel(dt_metrics = dt_metrics,
                               dt_average = dt_average,
                               cl_name = cl_name)

plot_dose_response_sa_qc_panel(dt_metrics = dt_metrics,
                               dt_average = dt_average,
                               cl_name = cl_name,
                               d_names = d_names)

Visualization for the quality control of the fitting for single-agent data

Description

Visualization for the quality control of the fitting for single-agent data

Usage

plot_fitting_acc(dt_assay, cl_name, normalization_type = "GR")

Arguments

dt_assay

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
cl_name

string cell line name to be plotted (Cell Line Name)
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")

Value

ggplot object containing panel with lollipop plots with r2 and rss values for each drug

Author(s)

Bartosz Czech [email protected]

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
cl_name <- dt_metrics[["CellLineName"]][1]

plot_fitting_acc(dt_assay = dt_metrics,
                 cl_name = cl_name,
                 normalization_type = "RV")

Plot data from a specific column

Description

Plot data from a specific column

Usage

plot_plate(dt_plate, column_name)

Arguments

dt_plate

The data table containing plate information
column_name

The name of the column to plot

Value

A named list of ggplot objects for each barcode and column

Author(s)

Bartosz Czech [email protected]

Examples

test_data <- data.table::data.table(
  WellColumn = rep(1:12, each = 8),
  WellRow = rep(LETTERS[1:8], times = 12),
  clid = "A",
  Gnumber = c(rep("untreated", 48), sample(1:5, size = 48, replace = TRUE)),
  Gnumber_2 = c(rep("untreated", 48), sample(1:5, size = 48, replace = TRUE)),
  Concentration = runif(96, min = 0, max = 100),
  ReadoutValue = runif(96, min = 0, max = 100),
  Barcode = rep(c("A", "B"), 48)
  )
plot_plate(test_data, "Gnumber")[["A"]]

Plot plate data (Heatmap + Dose Ranks + Smart Legend + Explicit QC)

Description

Plot plate data (Heatmap + Dose Ranks + Smart Legend + Explicit QC)

Usage

plot_plate_stack_info(
  dt_plate,
  ctrl_fail_threshold = 0.6,
  n_sd = 1,
  use_sd_threshold = TRUE,
  items_per_line = 6
)

Arguments

dt_plate

data.table. Input data containing plate layout, measurements, and standard gDR identifiers.
ctrl_fail_threshold

numeric (Default: 0.6). Flags control wells with a readout below Mean(Controls) * ctrl_fail_threshold.
n_sd

numeric (Default: 1). Sets the upper limit for flagging high signals using the formula Mean(Controls) + (n_sd * SD(Controls)).
use_sd_threshold

logical (Default: TRUE). If TRUE, uses the dynamic SD-based limit defined by n_sd. If FALSE, uses a fixed limit of Mean(Controls) * 1.1.
items_per_line

integer. Number of doses to show per line in the legend. Default 6.

Value

A named list of ggplot objects for each barcode

Author(s)

Bartosz Czech [email protected]

Examples

conc_series <- c(0, 0.001, 0.003, 0.01, 0.03, 0.1, 0.3, 1, 3, 10)

test_data <- data.table::data.table(
   WellColumn = sprintf("%02d", rep(1:12, each = 8)),
   WellRow = rep(LETTERS[1:8], times = 12),
   Barcode = rep(c("Plate_1", "Plate_2"), each = 48),
   clid = "CellLineA"
)

invisible(test_data[, Concentration := rep(rep(conc_series, length.out = 48), 2)])
invisible(test_data[, Gnumber := ifelse(Concentration == 0, "vehicle", "Drug_A")])
invisible(test_data[Gnumber == "vehicle", Concentration := 0])

invisible(test_data[, ReadoutValue := ifelse(Gnumber == "vehicle",
                                    rnorm(.N, 1000, 50),
                                    rnorm(.N, 500, 100))])
library(ggtext)
plots <- plot_plate_stack_info(test_data)
plots[[1]]

combo_data <- data.table::copy(test_data)
invisible(combo_data[, Barcode := "Plate_Combo"])
invisible(combo_data[, Concentration_2 := rep(rev(conc_series), length.out = .N)])
invisible(combo_data[, Gnumber_2 := ifelse(Concentration_2 == 0, "vehicle", "Drug_B")])
invisible(combo_data[Gnumber_2 == "vehicle", Concentration_2 := 0])

combo_plots <- plot_plate_stack_info(combo_data)
combo_plots[[1]]

Plot scatter with correlation

Description

Plot scatter with correlation

Usage

plot_scatter_with_corr(
  dt_response,
  dt_depmap,
  selected_feat,
  selected_feat_meta_col = NULL
)

Arguments

dt_response

data.table with experimental response data (rows are samples) for one metric outputted by one of functions: prep_dt_response_metric_sa, prep_dt_response_dose_sa, prep_dt_response_scores or prep_dt_response_metric_diff,
dt_depmap

data.table with dependent variables data loaded from DepMap - for one feature; (rows are samples, columns are features). outputted by prep_dt_depmap_feat
selected_feat

string with name of selected feature from dt_depmap
selected_feat_meta_col

string with the name of a feature column in DepMap (will be used as a plot title)

Value

ggplot object containing a scatter plot with correlation

Author(s)

Janina Smoła [email protected]

Examples

mae_prism <- gDRutils::get_synthetic_data("prism")
se_prism <- mae_prism[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_prism, "Metrics")
dt_response <- prep_dt_response_metric_sa(
  dt_metrics = dt_metrics,
  d_name = unique(dt_metrics[["DrugName"]])[1],
  normalization_type = "RV", metric = "x_mean")
feat_data_path <- system.file("depmap_data", package = "gDRtestData")
meta_data_path <- system.file("depmap_data/Model.csv.gz", package = "gDRtestData")
obj_depmap_feat <- prep_dt_depmap_feat(feat_data_path, meta_data_path,
                                       feature_set = "OmicsCNGene")
plot_scatter_with_corr(
  dt_response = dt_response,
  dt_depmap = obj_depmap_feat[["dt_depmap"]],
  selected_feat = names(obj_depmap_feat[["dt_depmap"]])[3])

Plot panel with scatter with correlation

Description

Plot panel with scatter with correlation

Usage

plot_scatter_with_corr_panel(
  dt_response,
  dt_depmap,
  selected_feats,
  selected_feat_meta_col = NULL,
  ncol = NULL
)

Arguments

dt_response

data.table with experimental response data (rows are samples) for one metric outputted by one of functions: prep_dt_response_metric_sa, prep_dt_response_dose_sa, prep_dt_response_scores or prep_dt_response_metric_diff,
dt_depmap

data.table with dependent variables data loaded from DepMap - for one feature; (rows are samples, columns are features). outputted by prep_dt_depmap_feat
selected_feats

character vector with names of selected features from dt_depmap
selected_feat_meta_col

string with the name of a feature column in DepMap (will be used as a plot title)
ncol

number of plot column in panel

Value

ggplot object containing panel of scatter plot with correlation for selected features

Author(s)

Janina Smoła [email protected]

Examples

mae_prism <- gDRutils::get_synthetic_data("prism")
se_prism <- mae_prism[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_prism, "Metrics")
dt_response <- prep_dt_response_metric_sa(
  dt_metrics = dt_metrics,
  d_name = unique(dt_metrics[["DrugName"]])[1],
  normalization_type = "RV", metric = "x_mean")
feat_data_path <- system.file("depmap_data", package = "gDRtestData")
meta_data_path <- system.file("depmap_data/Model.csv.gz", package = "gDRtestData")
obj_depmap_feat <- prep_dt_depmap_feat(feat_data_path, meta_data_path,
                                       feature_set = "OmicsCNGene")
plot_scatter_with_corr_panel(
  dt_response = dt_response,
  dt_depmap = obj_depmap_feat[["dt_depmap"]],
  selected_feats = names(obj_depmap_feat[["dt_depmap"]])[3:5])

Plot a single plate's stack info

Description

Generates the QC and dose-rank heatmap for a single plate subset. Handles all necessary data transformations (factor conversion, ranking, color mapping) internally.

Usage

plot_single_plate_stack_info(
  dt_subset,
  plate_id = NULL,
  drug_color_mapping = NULL,
  ctrl_fail_threshold = 0.6,
  n_sd = 1,
  use_sd_threshold = TRUE,
  items_per_line = 6
)

Arguments

dt_subset

data.table. Subset of the plate data corresponding to a single barcode/plate.
plate_id

character (Optional). The identifier (barcode) of the plate. If NULL, attempts to extract it from the data.
drug_color_mapping

character (Optional). Named vector of color codes. If NULL, generated automatically.
ctrl_fail_threshold

numeric (Default: 0.6). Flags control wells with a readout below Mean(Controls) * ctrl_fail_threshold.
n_sd

numeric (Default: 1). Sets the upper limit for flagging high signals using the formula Mean(Controls) + (n_sd * SD(Controls)).
use_sd_threshold

logical (Default: TRUE). If TRUE, uses the dynamic SD-based limit defined by n_sd. If FALSE, uses a fixed limit of Mean(Controls) * 1.1.
items_per_line

integer. Number of doses to show per line in the legend. Default 6.

Value

A ggplot object.

Author(s)

Bartosz Czech [email protected]

Examples

conc_series <- c(0, 0.001, 0.003, 0.01, 0.03, 0.1, 0.3, 1, 3, 10)

test_data <- data.table::data.table(
   WellColumn = rep(1:12, each = 8),
   WellRow = rep(LETTERS[1:8], times = 12),
   clid = "CellLineA",
   Barcode = "Plate_1"
)

invisible(test_data[, Concentration := rep(conc_series, length.out = .N)])
invisible(test_data[, Gnumber := ifelse(Concentration == 0, "vehicle", "Drug_A")])
invisible(test_data[Gnumber == "vehicle", Concentration := 0])

invisible(test_data[, ReadoutValue := ifelse(Gnumber == "vehicle",
                                    rnorm(.N, 1000, 50),
                                    rnorm(.N, 1000 * (1 / (1 + Concentration)), 50))])

library(ggtext)
plot_single_plate_stack_info(test_data)

invisible(test_data[, Concentration_2 := rep(rev(conc_series), length.out = .N)])
invisible(test_data[, Gnumber_2 := ifelse(Concentration_2 == 0, "vehicle", "Drug_B")])

invisible(test_data[Gnumber_2 == "vehicle", Concentration_2 := 0])

plot_single_plate_stack_info(test_data)

Plot a violin plot for normalized or averaged data (single-agent and combo) to control the quality of the data

Description

Plot a violin plot for normalized or averaged data (single-agent and combo) to control the quality of the data

Usage

plot_var_distribution_qc(
  dt_assay,
  cl_name,
  metric = "x",
  normalization_type = "GR"
)

Arguments

dt_assay

data.table representing data from the assay, outputted by gDRutils::convert_se_assay_to_dt(se, <assay_name>) for assay_name like Normalized or Averaged and SummarizedExperiment with chosen data type: single-agent or combo
cl_name

string cell line name to be plotted (Cell Line Name)
metric

string with variable name to be plotted; it has to be in dt_assay
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")

Value

ggplot object containing plot with violin for each drug

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_norm <- gDRutils::convert_se_assay_to_dt(se, "Normalized")
cl_name <- dt_norm[["CellLineName"]][1]

plot_var_distribution_qc(dt_assay = dt_norm,
                         cl_name = cl_name)

dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
plot_var_distribution_qc(dt_assay = dt_average,
                         cl_name = cl_name,
                         normalization_type = "RV")

plot_var_distribution_qc(dt_assay = dt_average,
                         cl_name = cl_name,
                         metric = "x_std",
                         normalization_type = "RV")

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]

dt_norm <- gDRutils::convert_se_assay_to_dt(se, "Normalized")
cl_name <- dt_norm[["CellLineName"]][1]

plot_var_distribution_qc(dt_assay = dt_norm,
                         cl_name = cl_name)

dt_average <- gDRutils::convert_se_assay_to_dt(se, "Averaged")
plot_var_distribution_qc(dt_assay = dt_average,
                         cl_name = cl_name,
                         normalization_type = "RV")

Lollipop plot for metric single-agent data to control quality of the data

Description

Lollipop plot for metric single-agent data to control quality of the data

Usage

plot_var_stat_qc(
  dt_assay,
  cl_name,
  metric = "r2",
  normalization_type = "GR",
  with_table = FALSE
)

Arguments

dt_assay

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt(se, "Metrics") and single-agent SummarizedExperiment
cl_name

string cell line name to be plotted (Cell Line Name)
metric

string with variable name to be plotted; it has to be in dt_assay
normalization_type

string with normalization_types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
with_table

logical whether table with metric values should be shown next to the plot

Value

ggplot object containing lollipop plot with stat value for each drug

Author(s)

Bartosz Czech [email protected]

Examples

mae <- gDRutils::get_synthetic_data("small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]

dt_metrics <- gDRutils::convert_se_assay_to_dt(se, "Metrics")
cl_name <- dt_metrics[["CellLineName"]][1]

plot_var_stat_qc(dt_assay = dt_metrics,
                 cl_name = cl_name)

plot_var_stat_qc(dt_assay = dt_metrics,
                 cl_name = cl_name,
                 metric = "r2",
                 normalization_type = "RV")

plot_var_stat_qc(dt_assay = dt_metrics,
                 cl_name = cl_name,
                 metric = "x_AOC",
                 normalization_type = "RV",
                 with_table = TRUE)

Volcano plot with association

Description

Volcano plot with association

Usage

plot_volcano_assoc(
  dt_assoc,
  selected_feat_meta_col,
  selected_metric,
  condition_info = NULL,
  alpha = 0.05,
  named_p_top = 10,
  max_N = NULL
)

Arguments

dt_assoc

data.table with the calculated linear association between DepMap and metrics outputted by calc_assoc
selected_feat_meta_col

string describing the name of the associated feature/metadata from DepMap
selected_metric

string describing the name of the selected metric used the association calculation
condition_info

string describing experiment condition (preferred: "DrugName"_"Gnumber"_"drug_moa"_"Duration")
alpha

numeric cutoff to identify statistically significant correlations
named_p_top

numeric value for p-top statistically significant correlations to be labeled on the plot
max_N

numeric value for limit the maximum number of non-statistically significant points to plot; for default NULL all points will be plotted.

Value

ggplot object containing a volcano plot with association

Examples

Y <- matrix(seq(0.5, 2, length.out = 50), nrow = 50,
            dimnames = list(sprintf("row_%s", 1:50), "met_1"))
X <- matrix(
  withr::with_seed(42, sample(c(NA, seq(0.35, 23.5, 1.25)),
                              size = 50*20, replace = TRUE)),
  nrow = 50,
  dimnames = list(sprintf("row_%s", 1:50), sprintf("feat_%s", 1:20)))
tab_assoc <- calc_assoc(X, Y)
plot_volcano_assoc(tab_assoc,
                   selected_feat_meta_col = "feat_XY",
                   selected_metric = "met_RV")

Plot panel with volcano plot and according to the data type - scatter plots or box plots

Description

Plot panel with volcano plot and according to the data type - scatter plots or box plots

Usage

plot_volcano_assoc_panel(
  dt_response,
  dt_depmap,
  selected_metric,
  selected_feat_meta_col
)

Arguments

dt_response

data.table with the experimental response data (rows are samples) for one metric outputted by one of functions: prep_dt_response_metric_sa, prep_dt_response_dose_sa, prep_dt_response_scores or prep_dt_response_metric_diff, must have at least a column with CellLineName and a numeric column with metric values.
dt_depmap

data.table with dependent variables data loaded from DepMap where rows are samples, columns are features/metadata levels; one of: data for one feature outputted by prep_dt_depmap_feat or data or data for one metadata outputted by prep_dt_depmap_meta
selected_metric

string name of the metric in dt_response
selected_feat_meta_col

string with name of selected feature from dt_depmap or the name of the selected metadata from dt_depmap - respectively

Value

A named list with elements:

  • assoc_data table with association data

  • panel ggplot object containing a panel with volcano plot and depending on data type: a scatter plots with correlation for top 4 variables or boxplots for variable levels

Author(s)

Janina Smoła [email protected]

Examples

mae_prism <- gDRutils::get_synthetic_data("prism")
se_prism <- mae_prism[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_prism, "Metrics")
dt_response <- prep_dt_response_metric_sa(
  dt_metrics = dt_metrics,
  d_name = unique(dt_metrics[["DrugName"]])[1],
  normalization_type = "RV", metric = "x_mean")
feat_data_path <- system.file("depmap_data", package = "gDRtestData")
meta_data_path <- system.file("depmap_data/Model.csv.gz", package = "gDRtestData")
obj_depmap_feat <- prep_dt_depmap_feat(feat_data_path, meta_data_path,
                                       feature_set = "OmicsCNGene")
plot_volcano_assoc_panel(
  dt_response = dt_response,
  dt_depmap = obj_depmap_feat[["dt_depmap"]],
  selected_metric = "RV_gDR_x_mean",
  selected_feat_meta_col = obj_depmap_feat[["selected_feat_meta_col"]])

Prep table with calculated linear associations

Description

Prep table with calculated linear associations

Usage

prep_dt_assoc(dt_response, dt_depmap, selected_feat_meta_col = NULL)

Arguments

dt_response

data.table with experimental response data (rows are samples) for one metric
dt_depmap

data.table with dependent variables data load from DepMap. (rows are samples, columns are features or meta); outputted by one of prep_dt_depmap_feat or prep_dt_depmap_meta
selected_feat_meta_col

string name of feature/meta column in DepMap

Value

A named list with elements, that may be input to plot_volcano_assoc

  • dt_assoc data.table with calculated association values between feature/meta of DepMap and selected metric,

  • condition_info string describing experiment condition (drugs),

  • selected_feat_meta_col string name of feature/meta.

Author(s)

Janina Smoła [email protected]

Examples

mae_prism <- gDRutils::get_synthetic_data("prism")
se_prism <- mae_prism[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se_prism, "Metrics")
dt_response <- prep_dt_response_metric_sa(
  dt_metrics = dt_metrics,
  d_name = unique(dt_metrics[["DrugName"]])[1],
  normalization_type = "RV", metric = "x_mean")
feat_data_path <- system.file("depmap_data", package = "gDRtestData")
meta_data_path <- system.file("depmap_data/Model.csv.gz", package = "gDRtestData")
obj_depmap_feat <- prep_dt_depmap_feat(feat_data_path, meta_data_path,
                                       feature_set = "OmicsCNGene")
prep_dt_assoc(
  dt_response = dt_response,
  dt_depmap = obj_depmap_feat[["dt_depmap"]],
  selected_feat_meta_col = obj_depmap_feat[["selected_feat_meta_col"]])

Prep table with metric values by doses for single-agent experiment

Description

Prep table with metric values by doses for single-agent experiment

Usage

prep_dt_response_dose_sa(
  dt_average,
  d_name,
  normalization_type = "RV",
  metric = "x",
  fit_source = "gDR"
)

Arguments

dt_average

data.table representing data from the Averaged assay, outputted by gDRutils::convert_se_assay_to_dt and SummarizedExperiment with chosen data type: single-agent or combo
d_name

string with drug name to be plotted (identifiers DrugName)
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "x" (value of "GR" or "RV" itself - respectively depending on normalization_type), or "x_std" (standard deviation)
fit_source

string source name for metrics

Value

data.table with selected metric, input to prep_dt_assoc

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix_small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]
dt_average <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Averaged")
d_name <- "drug_004"
dt_response <- prep_dt_response_dose_sa(dt_average, d_name)

Prep table with metric values for combination experiment

Description

Prep table with metric values for combination experiment

Usage

prep_dt_response_metric_diff(
  dt_metrics,
  d_name,
  d_name2,
  cellline1 = NULL,
  cellline2 = NULL,
  normalization_type = "RV",
  metric = "xc50",
  fit_source = "gDR",
  additional_cols = NULL
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt and combo SummarizedExperiment
d_name

string representing the drug name to be plotted (identifier DrugName). If set to NULL, the function will return a table for all available DrugName
d_name2

string representing the drug name to be plotted (identifier DrugName_2). If set to NULL, the function will return a table for all available DrugName_2
cellline1

string representing the first cell line name. If set to NULL, the function will return a table for all available cell lines.
cellline2

string representing the second cell line name. If set to NULL, the function will return a table for all available cell lines.
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max"), "x_mean" ("GR Mean" or "RV Mean") "ec50" ("GEC50" or "EC50") or "x_AOC_range" ("GR AOC within set range" or "RV AOC within set range")
fit_source

string source name for metrics
additional_cols

character vector with additional cols that should be included in the output

Value

data.table with selected metric, input to prep_dt_assoc

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix_small")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Metrics")
d_name <- "drug_004"
d_name2 <- "drug_026"
dt_response <-
  prep_dt_response_metric_diff(dt_metrics, d_name, d_name2,
  metric = c("xc50", "x_mean", "x_max"))

cellline1 <- "cellline_GB"
cellline2 <- "cellline_HB"

dt_response <-
  prep_dt_response_metric_diff(dt_metrics, d_name = NULL, d_name2 = NULL,
  cellline1, cellline2,
  metric = c("xc50", "x_mean", "x_max"))

Prep table with metric values for single-agent experiment

Description

Prep table with metric values for single-agent experiment

Usage

prep_dt_response_metric_sa(
  dt_metrics,
  d_name,
  normalization_type = "RV",
  metric = "xc50",
  fit_source = "gDR"
)

Arguments

dt_metrics

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt and single-agent SummarizedExperiment
d_name

string with drug name to be plotted (identifiers DrugName)
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean")
fit_source

string source name for metrics

Value

data.table with selected metric, input to prep_dt_assoc

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix_small")
se <- mae[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Metrics")
d_name <- "drug_004"
dt_response <- prep_dt_response_metric_sa(dt_metrics, d_name)
dt_response <-
  prep_dt_response_metric_sa(dt_metrics, d_name,
                             metric = c("xc50", "x_mean", "x_max"))

dt_averaged <- gDRutils::convert_se_assay_to_dt(se = se,
                                                assay_name = "Averaged")
dt_metrics_capped <-
  gDRutils::cap_assay_infinities(conc_assay_dt = dt_averaged,
                                 assay_dt = dt_metrics,
                                 experiment_name = gDRutils::get_supported_experiments("sa"),
                                 capping_fold = 5)
d_name <- "drug_026"
dt_response <- prep_dt_response_metric_sa(dt_metrics, d_name)
dt_response <-
  prep_dt_response_metric_sa(dt_metrics, d_name,
                             metric = c("xc50", "x_mean", "x_max"))

Prep table with metric values for combination experiment

Description

Prep table with metric values for combination experiment

Usage

prep_dt_response_scores(
  dt_scores,
  d_name,
  d_name2,
  normalization_type = "RV",
  metric = "hsa_score",
  fit_source = "gDR"
)

Arguments

dt_scores

data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt and combo SummarizedExperiment
d_name

string with drug name to be plotted (identifiers DrugName)
d_name2

string with drug name to be plotted (identifiers DrugName_2)
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the combo metric; one of: "hsa_score"("Bliss Excess GR" or "Bliss Excess RV" - respectively depending on normalization_type), "bliss_score" ("Bliss Score GR" or "Bliss Score RV")
fit_source

string source name for metrics

Value

data.table with selected metric, input to prep_dt_assoc

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix_small")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_scores <- gDRutils::convert_se_assay_to_dt(se = se,
                                              assay_name = "scores")
d_name <- "drug_004"
d_name2 <- "drug_026"
dt_response <- prep_dt_response_scores(dt_scores, d_name, d_name2)
dt_response <-
  prep_dt_response_scores(dt_scores, d_name, d_name2,
                          metric = c("hsa_score", "bliss_score"))

Prep matrix with metric value based on the Metrics assay

Description

Prep matrix with metric value based on the Metrics assay

Usage

prep_pheatmap_matrix(
  dt_response,
  normalization_type = "GR",
  metric = "xc50",
  fit_source = "gDR",
  experiment_type = gDRutils::get_supported_experiments("sa")
)

Arguments

dt_response

data.table representing data from the Metrics assay, outputted by gDRutils::convert_se_assay_to_dt and single-agent SummarizedExperiment or data.table representing data from the scores assay, outputted by gDRutils::convert_se_assay_to_dt and combo SummarizedExperiment
normalization_type

string with normalization types to be selected one of: "GR" ("GRvalue") or "RV" ("RelativeViability")
metric

string name of the metric; one of: "xc50" ("GR50" or "IC50" - respectively depending on normalization_type), "x_max" ("GR Max" or "E Max") or "x_mean" ("GR Mean" or "RV Mean"); but the values from any numeric column can be displayed.
fit_source

string source name for metrics
experiment_type

string with experiment name; one of: "single-agent", "combination" or "co-dilution"

Value

matrix with values for selected metric with CellLinName in the rows and DrugName (or combination of DrugName and DrugName_2) in the columns

Author(s)

Janina Smoła [email protected]

Examples

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("sa")]]
dt_metrics <- gDRutils::convert_se_assay_to_dt(se = se,
                                               assay_name = "Metrics")

mat <- prep_pheatmap_matrix(dt_response = dt_metrics,
                            experiment_type = gDRutils::get_supported_experiments("sa"))

mae <- gDRutils::get_synthetic_data("combo_matrix")
se <- mae[[gDRutils::get_supported_experiments("combo")]]
dt_scores <- gDRutils::convert_se_assay_to_dt(se = se,
                                              assay_name = "scores")

mat <- prep_pheatmap_matrix(dt_response = dt_scores,
                            metric = "hsa_score",
                            normalization_type = "RV",
                            experiment_type = gDRutils::get_supported_experiments("combo"))

Round numbers to unique string

Description

Rounds a numeric vector to the minimum precision needed to ensure uniqueness

Usage

round_to_unique_string(num_vec, initial_digits = 4)

Arguments

num_vec

a numeric vector to be rounded and change into character
initial_digits

numeric value for number of decimal places to start rounding with

Value

a character vector of unique numeric strings.

Author(s)

Janina Smoła [email protected]

Examples

vec <- c(0.00000000, 0.00000256, 0.00001280, 0.00006400, 0.00032000,
         0.00160000, 0.00800000, 0.04000000, 0.20000000, 1.00000000)

round_to_unique_string(vec)