Benchmarking existing computational approaches for detecting cell-type-specific spatially variable genes (ctSVGs) in spatial transcriptomics data.
Scripts for preprocessing, analyzing, and visualizing real spatial transcriptomics datasets.
real/fig4_rotate_sc.R: Rotate spatial coordinates and run ctSVG methods (single-cell spatial data).real/fig4_rotate_spot.R: Rotate spatial coordinates and run ctSVG methods (spot-level spatial data).real/real.datasets1.R: Run benchmarking on real spatial transcriptomics datasets (spot-level).real/real.datasets2.R: Run benchmarking on real datasets with spot subsetting.real/real.data3.sc.R: Run benchmarking on real datasets at single-cell spatial resolution.real/scability.R: Scalability evaluation on real datasets (runtime/memory under varying spot sizes).
real/preprocess_1.1.py: Preprocessing spatial datasets.real/preprocess_1.2_public_data.R: Preprocess public spatial transcriptomics datasets used in benchmarks.real/preprocess_1.3_lung_cancer.R: Preprocess lung cancer spatial transcriptomics dataset.real/preprocess_2_bounary.R: Perform boundary preprocessing required by spVC.real/preprocess_3_subset_spots.R: Preprocesses data for subset spots.
real/plot_fig2.R: Figure 2 — consistency analysis on real datasets.real/plot_fig4_rotate_sc.R: Figure 4 — rotation results (single-cell spatial data).real/plot_fig4_rotate_spot.R: Figure 4 — rotation results (spot-level spatial data).real/plot_fig5_time_mem.R: Figure 5 — runtime and memory benchmarking.real/plot_fig6_lung cancer.R: Figure 6 — lung cancer results.real/plot_fig7_mbm.R: Figure 7 — MBM dataset results.real/plot_fig8_summary.R: Figure 8 — summary of overall method performance.real/plot_figs1_datasets.R: Supplementary Fig S1 — overview of datasets used in the benchmark.real/plot_lung_figs.R: Lung cancer visualization results in supplementary figures.
real/utils/get_wide_pval.R: Helper function for reshaping p-value results.real/utils/real-bench.R: Real-data benchmark helpers.real/utils/real_expr_bench.R: Expression / p-value correlation benchmark helpers.real/utils/rotate_bench.R: Rotation benchmark helpers.real/utils/rotate_subset.R:Subsetting helper functions for rotation.
Scripts for generating and analyzing simulated datasets.
sim/1-runsim_sc.R: Run simulation pipeline for sc data and call ctSVG methods.sim/1-runsim_spot.R: Run simulation pipeline for spot data and call ctSVG methods.sim/1-runsim_noRCTD.R: Run simulation pipeline without RCTD.sim/fig3-permutation.R: Build permutation datasets and run ctSVG methods (for Fig 3).
sim/plot_fig3.R: Figure 3 — main simulation results.sim/plot_fig3f_permutation.R: Figure 3f — permutation results visualization.sim/plot_nodeconv.R: Plot deconvolution influence.sim/plot_Supplementary fig prop_fpr.R: Supplementary figure — plots the relationship between false positive rate (FPR) and cell-type proportion.sim/plot_Supplementary figs auc.R: Supplementary figures — AUC with different drop-outs.sim/pre_fig3B.R: Plot figure 3B utils.
sim/utils/generate_sc.R: Generate simulated single-cell reference data.sim/utils/generate_st.R: Generate simulated spatial transcriptomics datasets.sim/utils/run_analysis_for_pattern.R: Runs ctSVG analysis for simulated data.sim/utils/run_analysis_for_pattern_sc.R: Runs ctSVG analysis for simulated single-cell level spatial data.sim/utils/run_analysis_for_pattern_sp_noRCTD.R: Runs ctSVG analysis for simulated data (without deconvolution).sim/utils/calc_false_positive_rate.R: Compute false positive rate metrics.sim/utils/sim-bench-sc.R: Simulation benchmark helpers (sc and sp spatial data).sim/utils/sim-bench-nodeconv.R: Simulation benchmark helpers (without deconvolution).
my_theme.R: Custom ggplot theme and style settings used across most figures.