BEAR-GRN: Benchmarking, Evaluation and Assessment Resource for GRNs

This directory contains the data required to reproduce the benchmarking results as reported in BEAR-GRN. This data is also required for benchmarking new GRN inference methods utilizing multiomics. INFERRED.GRNS - contains the inferred GRNs from Eight GRN inference methods for eight dataset used in BEAR-GRN GROUND.TRUTHS - contain the all primary ground truths filtered for expressed genes in each dataset STABILITY_GRNS - contain the inferred GRNs from mESC subsamples created for stability measument for Eight GRN inference methods INPUT.DATA - Contains processed multiomic data used for GRN inference (Gene by Cell and Peak by Cell matrix) INPUT.DATA.STABILITY - Contains rds file of subsampled scRNA and scATAC data and a script that can be used to generate matrix and create subsamples directories using RDS files GENOME_AND_ANNOTATION - contains the genome, annotation, motif annotation, PWM files used by these inference algorithms for reproducibility GROUND.TRUTHS.KO - contain the Knock out ground truth filtered for expressed genes in each dataset GROUND.TRUTHS.UNION - contain the union of Chipseq and Knock out ground truth filtered for expressed genes in each dataset GROUND.TRUTHS.INTERSECTION - contain the the intersection of Chipseq and Knock out ground truth filtered for expressed genes in each dataset CORE_GROUND_TRUTH - contain the core ground truth created from all human ground truth datasets used in this study; filtered for expressed genes in each dataset CELL.TYPE.EXCLUSIVE.GROUND.TRUTH - contain the cell type exclusive ground truth created for all human ground truth datasets used in this study; filtered for expressed genes in each dataset Note: The ground truth file name are same in each GT dir, this is intentional to support the GT in BEARGRN, as the name of GTs are hardcoded in BEARGRN.