Package: spiDE 0.99.37

Dharmesh D. Bhuva

spiDE: Context-specific, neighbourhood-dependent differential expression

spiDE identifies context-specific, neighbourhood-dependent differential expression in spatial transcriptomics data. Within an index cell type, it asks whether a gene's expression changes with the local density (niche) of another cell type within patients, and whether that dependence differs between two conditions. Niche covariates are Gaussian kernel density estimates of each cell type. For every index cell type a per-gene negative binomial GLM with a patient intercept is fitted with SpaNorm's Newton solver; its niche slopes are then tested across patients, either from each patient's own slopes with weighted limma (the slopes engine) or with a patient-clustered CR2 sandwich (the sandwich engine). The pooled test asks whether a slope is consistently non-zero; the condition-specific test whether it differs between conditions.

Authors:Dharmesh D. Bhuva [aut, cre], Chin Wee Tan [aut]

spiDE_0.99.37.tar.gz
spiDE_0.99.37.zip(r-4.7-any)spiDE_0.99.37.zip(r-4.6-any)spiDE_0.99.37.zip(r-4.5-any)
spiDE_0.99.37.tgz(r-4.6-any)spiDE_0.99.37.tgz(r-4.5-any)
spiDE_0.99.37.tar.gz(r-4.7-any)spiDE_0.99.37.tar.gz(r-4.6-any)
spiDE_0.99.37.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
spiDE/json (API)

# Install 'spiDE' in R:
install.packages('spiDE', repos = c('https://biocstaging.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/bhuvalab/spide/issues

Datasets:
  • toySpiDE - A small synthetic SpatialExperiment for spiDE examples

On CRAN:

Conda:

softwarespatialdifferentialexpressiongeneexpressiontranscriptomicsregressioncellbiology

4.53 score 7 scripts 21 exports 116 dependencies

Last updated from:612def34f3. Checks:1 WARNING, 9 OK. Indexed: yes.

TargetResultTimeFilesSyslog
bioc-checksWARNING220
linux-develOK396
source / vignettesOK327
linux-releaseOK415
macos-releaseOK333
macos-oldrelOK246
windows-develOK341
windows-releaseOK263
windows-oldrelOK355
wasm-releaseOK159

Exports:buildNichesfitSpiDEmergeNichespatientInterceptspatientSlopesplotCallMapplotNicheMapplotNicheResponseplotPatientEffectsplotPatientSlopesplotPValuesplotSpilloverplotTopTripletsplotTripletHeatmapresultsspiDEspiDEColoursspiGSEAtestNicheAbundancetestSpiDEtheme_spiDE

Dependencies:abindaskpassassortheadbeachmatBHBiobaseBiocFileCacheBiocGenericsBiocNeighborsBiocParallelBiocSingularbitbit64blobblustercachemcliclustercodetoolscpp11curlDBIdbplyrDelayedArraydeldirdotCall64dplyrdqrngedgeRfarverfastmapfilelockformatRfutile.loggerfutile.optionsfuturefuture.applygenericsGenomicRangesggplot2globalsgluegoftestgtablehttr2igraphIRangesirlbaisobandlabelinglambda.rlatticelifecyclelimmalistenvmagickmagrittrMatrixMatrixGenericsmatrixStatsmemoisemetapodnlmeopensslparallellypillarpkgconfigpolyclipprogressrpurrrR6RColorBrewerRcppRcppEigenrjsonrlangRSQLitersvdS4ArraysS4VectorsS7ScaledMatrixscalesscranscuttleSeqinfoSeuratObjectSingleCellExperimentsitmosnowspspamSpaNormSparseArraySpatialExperimentspatstat.dataspatstat.explorespatstat.geomspatstat.randomspatstat.sparsespatstat.univarspatstat.utilsstatmodstringistringrSummarizedExperimentsystensortibbletidyrtidyselectutf8vctrsviridisLitewithrXVector

Designing a study with spiDE
How much data does a study need? | The benchmark | Reading the minimum detectable effect | Patients and cells per patient | The condition-specific test | Small designs: 5 to 7 patients per condition | False discoveries to expect | Practical recommendations | Session information
Rendered fromspiDE-design.Rmdusingknitr::rmarkdown

Last update: 2026-10-22
Started: 2026-10-22

Getting started with spiDE
spiDE | Load the data | Build the niches | Choose a bandwidth | Fit the model | Test the slopes | The pooled test | The condition-specific test | Look at the patients behind a result | Test gene sets | A different question: niche abundance | Plots | Where the calls are | The cells behind a slope | Check for spillover | Diagnostics | Practicalities | Session information
Rendered fromspiDE.Rmdusingknitr::rmarkdown

Last update: 2026-10-22
Started: 2026-07-10

The spiDE model
The question | What kinds of effects can spiDE find? | both groups spread evenly over the same range of patient mean densities | per panel and patient: the mean niche density m, the level a at that | density, and the within-patient slope b | each patient's own least-squares line over the central 90% of its | densities, and the patient's mean
Rendered fromspiDE-model.Rmdusingknitr::rmarkdown

Last update: 2026-10-22
Started: 2026-07-10

Validation: calibration and accuracy
Can we trust the calls? | How calibration is checked | The cohorts | The slopes engine (the default) | The sandwich engine | Accuracy and power, with a known truth | Sequencing depth | Gene sets | The between-patient test | What has not been measured | Session information
Rendered fromspiDE-calibration.Rmdusingknitr::rmarkdown

Last update: 2026-10-22
Started: 2026-08-04

Readme and manuals

Help Manual

Help pageTopics
Build niche covariate matricesbuildNiches buildNiches,ANY-method
Fit spiDE's per-patient niche modelsfitSpiDE fitSpiDE,SpatialExperiment-method
Merge niche cell-type columns into groupsmergeNiches mergeNiches,ANY-method
Each patient's interceptpatientIntercepts patientIntercepts,SpiDEFit-method patientIntercepts,SpiDEResults-method
Each patient's niche slopespatientSlopes patientSlopes,SpiDEFit-method patientSlopes,SpiDEResults-method
Where the calls areplotCallMap
The niche covariate in tissueplotNicheMap
Expression against niche density: the cells behind a slopeplotNicheResponse
What the patient intercepts, or slopes, captureplotPatientEffects
Each patient's niche slope for a tripletplotPatientSlopes
p-values by test and expression bandplotPValues
Are the calls the niche type's own markers?plotSpillover
The strongest triplets, or those of chosen genesplotTopTriplets
Niche-response profiles of genes or gene setsplotTripletHeatmap
Extract spiDE's test resultsresults results,SpiDEResults-method
Run spiDE end to endspiDE spiDE,SpatialExperiment-method
The spiDE coloursspiDEColours
The fitted per-patient niche models of spiDE$,SpiDEFit-method SpiDEFit-class
The results of spiDE's niche tests$,SpiDEResults-method SpiDEResults-class
Gene-set tests on spiDE's niche resultsspiGSEA spiGSEA,SpiDEResults-method
Test whether patients with more of a niche cell type express genes differentlytestNicheAbundance testNicheAbundance,ANY-method
Test niche-dependent differential expressiontestSpiDE testSpiDE,SpiDEFit-method
The spiDE plot themetheme_spiDE
A small synthetic SpatialExperiment for spiDE examplestoySpiDE