METANet: TF network maps for 36 human tissues

This repository accompanies the METANet paper and contains all data and software needed to reproduce or apply the METANet framework across 36 human GTEx tissues. Contents: METANet_compendium.zip — Predicted TF gene regulatory networks for 36 human GTEx tissues. Includes: METANets: TF regulatory networks predicted by METANet, with edge weights representing predicted probabilities of TF binding to the cis-regulatory elements of target genes.Tissue-specific METANets: METANets filtered for tissue-specific edges using an IQR-based method.Other networks: Benchmark and ablated networks evaluated in the study, including Marbach et al. (2016) networks, PANDA (Sonawane et al. 2017), tissue-specific PANDA, ETANet, META-LASSO, META-BART, META-TAEx, META-TSEx, METANet-5, and METANet-20. XGBoost_models: Includes the trained XGBoost models (JSON format) and TF-disjoint cross-validation fold assignments used to train METANets. METANet-1.0.1-GitHub.zip — Source code release of the METANet framework (mirrored from GitHub). Contains the full pipeline for feature construction, model training, network prediction, and evaluation.