ENCODE Releases 2,300+ AI Models for Gene Regulation on Hugging Face

anshulkundaje · x · 2026-08-05

Accompanying the ENCODE 4 project, Stanford's Kundaje Lab released the ENCODE BPNet models collection on Hugging Face. The collection includes deep learning models trained on 2,339 TF ChIP-seq datasets across 788 human transcription factors.

These models are also integrated into the UCSC Genome Browser and the ENCODE Portal, providing researchers with accessible tools to analyze and visualize gene regulatory mechanisms.

Related event: Stanford Releases ENCODE GRAMMAR: Thousands of AI Models to Decode Human Genome(12 posts)→

Original post →

More from Research

Research channel →