Reproduce “Beyond Action Entropy: Quotient-Space Exploration for Genome-Scale Metabolic Model Repair” 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

Xuan Gong, Hanbo Huang, Wenbin Dai, Jing Wang, Lei Bai, Xiang Xiao, Weishu Zhao, Shiyu Liang

QuotientPO improves metabolic model repair by exploring canonical repair cores rather than raw reaction-edit lists. It removes edits that can be deleted without changing phenotype predictions, then rewards less-crowded cores. On 2,212 held-out genome-scale models, it raises Success@32 from 17.93% to 20.10% and Success Cores@32 from 0.50 to 0.57 versus Vanilla GRPO. These verifier-based results do not establish in-vivo mechanisms.

Summary by alphaXiv.

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