Stanford Releases ENCODE GRAMMAR: 3,865 Deep Learning Models to Decode the Human Genome
anshulkundaje · x · 2026-08-05
Stanford's Kundaje Lab, in collaboration with the NIH ENCODE consortium, released ENCODE GRAMMAR, a massive deep learning model resource designed to decode the DNA sequence logic of regulatory elements in the human genome.
This multi-year effort includes 3,865 experiment-specific deep learning model sets and sequence annotations, covering key dimensions like transcription factor binding and chromatin accessibility.
Open Access & Tooling
All resources are fully open-source and accessible via:
- The ENCODE Portal
- Hugging Face
- UCSC Track Hub
- ENCODE Motif Compendium
The team also published a 5-minute quickstart guide to help users load observed accessibility profiles and model predictions into the WashU Epigenome Browser. Deeper dives into model interpretation and genetic variant effect prediction will be released weekly.
More from Research
- Harvard, MIT, Stanford Launch MatrAIx: 8.3B Persona Agents for AI Evaluation — BrihiJ · 2026-08-06
- Dongfang Starlink Launches AI Hyperspectral Satellites, Plans 2M Lunar Robots for Data Center — 新智元 · 2026-08-06
- USTC Study: AI Agents Achieve Only 3.3% Zero-Intervention Execution in Robot Labs — 新智元 · 2026-08-06
- 23-Year-Old OpenAI Alignment Researcher Joins Brain-Computer Startup Conduit — 新智元 · 2026-08-06
- New OGM Neural Network Architecture for Disease Risk Prediction — anshulkundaje · 2026-08-06
- CoCoEvolve: A New Framework for Cross-Representation Consistency in Charts, Tables, and Code — Xuehang Guo · 2026-08-06