In setting_files_for_grnboost2 there are setting files for running GRNBoost2 on scRNAseq and BIOS data
GRNBoost2.ipynb: Prepare the input files for BEELINE, and examine the GRNBoost2 results for scRNAseq and for BIOS data
rho_comparison_lowlyexpressed.R: explores the differences between Spearman correlation and Rho propensity specially for very lowly expressed genes
scorpius_and_slingshot_clean.R: calculates the pseudotime ordering for Oelen v2 classical monocytes, using SCORPIUS and Slingshot algorithms
scvelo_analysis_dm.py: runs RNA velocity analysis on Oelen v3 dataset classical monocytes after creating loom files using velocyto to get both spliced and unspliced gene count matrices
compare_cell_classification.ipynb: compares the aximuth cell type classification with the marker gene cell type classification in Oelen v2 and v3 dataset for untreated cells
Metacell calculation and evaluation files are all in the directory metacell: metacell_per_sample_original_algorithm.R: calculates metacells based on original algorithm (implemented in the metacell R package) metacells_from_leiden.R: calculates metacells based on grouping from leiden clustering create_genesets.R: split all genes expressed in Oelen v3 dataset, Monocytes, into different expression bins for treshold-dependent evaluation with BLUEPRINT metacell_general_correlation_tp.R: calculates correlation from metacells (original or leiden) for different expression tresholds from create_genesets.R for comparison with BLUEPRINT single_cell_correlation_tp.R: calculates correlation from single cell dataset for different expression tresholds from create_genesets.R for comparison with BLUEPRINT eval_blueprint_genesets.R: compares correlation from BLUEPRINT with correlation from metacells/single cell for different expression tresholds, using correlation vlaues from metacell_general_correlation_tp.R and single_cell_correlation_tp.R plot_overview_metacell.R: visualize outputs from metacell evaluation in one plot