Reproduce “AutoNovo: Agentic Program Search Discovers Transferable Nucleic Acid Design Algorithms” 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

Joel Shor, Subhashini Venugopalan

AutoNovo uses an AI system to improve DNA-design algorithms rather than validate sequences experimentally. It revises designer code and evaluates designs with models of K562 chromatin accessibility. Across three design tasks and six cross-model transfer directions, its algorithms achieved the best observed mean in each; these are in-silico scores, not evidence of wet-lab validity, and different algorithms led different tasks.

Summary by alphaXiv.

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