--- title: "Interactive Single-Cell Simulation Benchmarking with scSimEvalStudio" output: rmarkdown::html_vignette: toc: true toc_depth: 3 vignette: > %\VignetteIndexEntry{Interactive Single-Cell Simulation Benchmarking with scSimEvalStudio} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8} --- ```{r setup, include = FALSE} knitr::opts_chunk$set( collapse = TRUE, comment = "#>" ) ``` # Introduction **`scSimEvalStudio`** provides the interactive Shiny graphical studio companion for **`scSimEval`**. While `scSimEval` focuses cleanly on Bioconductor programmatic evaluation across `SingleCellExperiment` objects and reproducibility pipelines, `scSimEvalStudio` exposes point-and-click dataset uploads, real-time evaluation progress logs, pre-saved `.rds` exploration, interactive zoomable plots, and Excel/figure exports. # Launching the Shiny Studio After installing both `scSimEval` and `scSimEvalStudio`, launch the interactive studio from R: ```{r launch-snippet, eval = FALSE} library(scSimEvalStudio) launch_scSimEval_app() ``` By default, `launch_scSimEval_app()` configures `shiny.maxRequestSize` to 10 GiB (`10 * 1024^3`) so large multiomics `.rds` archives can be loaded locally. # Verifying Application Initialization In automated or non-interactive environments, you can confirm that the bundled Shiny application initializes cleanly and inspects the bundled demonstration benchmark dataset: ```{r verify-app} library(scSimEvalStudio) app_dir <- system.file("shiny", "scSimEvalApp", package = "scSimEvalStudio") dir.exists(app_dir) app <- shiny::shinyAppDir(app_dir) class(app) demo_file <- file.path(app_dir, "data", "demo_benchmark_data.rds") demo_obj <- readRDS(demo_file) names(demo_obj) head(demo_obj$benchmark_summary_table[, c("Method", "Category", "Metric", "Score")], 6) ``` # Bioconductor Ecosystem Integration `scSimEvalStudio` accepts `SingleCellExperiment` objects (including multi-modal `SingleCellExperiment` objects with `altExp` assays and `colData` annotations `cell_type` and `batch`) evaluated directly by `scSimEval::evaluate_simulation_accuracy()` and `scSimEval::evaluate_multiomics_accuracy()`. # Session Information ```{r session-info} sessionInfo() ```