🧬 Generative modeling of regulatory DNA sequences with diffusion probabilistic models 💨

Elucidating the Utility of Genomic Elements with Neural Nets

CREsted is a Python package for training sequence-based deep learning models on scATAC-seq data, for capturing enhancer code and for designing cell type-specific sequences.

surrogate quantitative interpretability for deepnets

Robust and efficient analysis of single-cell perturbation studies

Genomic sequence preprocessing toolkit

Data-driven design of context-specific regulatory elements

A unified framework for discovering, analyzing, integrating, and visualizing regulatory motifs and transcription factor binding sites across bulk, single-cell, and long-read sequencing modalities.

lsgkm+gkmexplain with regression functionality. Builds off kundajelab/lsgkm (which has gkmexplain), which in turn builds off Dongwon-Lee/lsgkm (the original lsgkm repo)

Interpreting sequence-to-function machine learning models

A set of tutorials for how to use all the tools in ML4GLand

Dual Threshold Optimization compares two ranked lists of features (e.g. genes) to determine the rank threshold for each list that minimizes the hypergeometric p-value of the overlap of features. It then calculates a permutation based empirical p-value and an FDR

A curated list of regulatory genomics papers and resources.

Integrative framework combining TF footprinting with genome-wide association analyses to identify causal noncoding variants and elucidate their regulatory mechanisms in gene regulation

Threshold and p-value computations for Position Weight Matrices

Prokaryote Gene Regulatory Network (ProGRN) Inference Pipeline

squid repository for manuscript analysis

Deep learning model for non-coding regulatory variants

Repository documenting applications of the ML4GLand suite on published datasets

Datasets for benchmarking, testing and developing in EUGENe