Reproduce “Genome-Factory: A Library for Tuning, Deploying, and Interpreting Genomic Foundation Models” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Weimin Wu, Xuefeng Song, Yibo Wen, Qinjie Lin, Zhihan Zhou, Jerry Yao-Chieh Hu, Zhong Wang, Han Liu

Genome-Factory is a Python library that connects genomic sequence preparation, model loading and tuning, inference, benchmarking, and representation analysis in one workflow. In a DNABERT-2 COVID-variant task, adapter tuning used 0.20 million trainable parameters and 2.09 GB peak GPU memory, at 124.30 thousand tokens/second. These measurements describe one setup, not universal performance or validated biological discovery.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.