Reproduce “MEMIG: Self-evolving Memory Construction for Antibody Optimization” 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

Yunhui Jang, Zach Wang, Jai Doshi, Kangway Chuang, Sidney Lisanza, Taylor Joren

MEMIG helps agents optimize antibody sequences by storing oracle-measured mutation effects under identifiers for target property, germline, antibody region, and mutation, and by recording pairwise interactions. In an in-silico binding task, it achieved the highest Pass@1 among non-structural methods in 9 of 11 clusters with GLM-5.3-Flash. Passing required predicted pTM > 0.8 and ipTM > 0.7; experimental binding was not tested.

Summary by alphaXiv.

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