genomicsxai

genomicsxai / alphagenome-pytorch

Public

AlphaGenome PyTorch port

70
10
100% credibility
Found Mar 12, 2026 at 53 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

PyTorch reimplementation of DeepMind's AlphaGenome model for predicting genomic tracks from long DNA sequences.

How It Works

1
🔍 Discover the genomic predictor

You hear about a smart tool that reads DNA sequences and predicts important biological signals like accessibility and gene activity.

2
📦 Get it set up quickly

You add it to your computer with a simple install command, no hassle.

3
🧠 Load the ready model

You bring in the pre-trained brain that already understands human and mouse genomes.

4
🧬 See predictions instantly

Give it a DNA snippet and watch it reveal hundreds of tracks showing where genes turn on, chromatin opens, and proteins bind.

5
🎯 Tune it to your data

Feed your own experiments to customize predictions for your specific cells or conditions.

✨ Unlock biology insights

Your tailored predictions help answer questions about gene regulation and disease.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 53 to 70 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is alphagenome-pytorch?

alphagenome-pytorch is a Python PyTorch port of DeepMind's AlphaGenome DNA sequence model. It predicts hundreds of genomic tracks—like chromatin accessibility, TF binding, RNA expression, and 3D contacts—at single base-pair or 128bp resolution from one-hot encoded sequences up to 1Mbp. Load pretrained weights from Hugging Face and run inference or fine-tune via simple pip install and a few lines of code.

Why is it gaining traction?

It delivers numerical parity with the original JAX model, validated through forward-pass comparisons and gradient checks, so predictions match exactly. Built-in fine-tuning CLI handles LoRA, linear probing, or full training on BigWig data with multi-GPU support, plus notebooks for variant scoring and full-chromosome predictions as BigWig outputs. PyTorch-native means it slots into existing pipelines without JAX overhead.

Who should use this?

Genomics researchers fine-tuning on custom ATAC-seq or RNA-seq BigWigs for cell-type models. Bioinformaticians scoring variants or running in silico mutagenesis without wrestling JAX checkpoints. PyTorch teams prototyping genomic apps, like regulatory element predictors.

Verdict

Grab it if JAX is a barrier—strong docs, example notebooks, and validation scripts make this credible despite 48 stars and 1.0% score. Early but production-ready for experiments; verify outputs before pipelines.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.