Reproduce “From Sequence Uncertainty to Phenotype Hypotheses: Uncertainty-Conditioned Inference over Biomedical Knowledge Graphs” 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

Huixin Zhan, Jason H. Moore

The authors turn calibrated sequence-model uncertainty into a distribution over valid gene-to-phenotype graph routes, scoring routes and summing their probabilities by phenotype. In AlzKB, uncertainty correlates with phenotype-group entropy (Spearman’s ρ=0.951), versus 0.582 at fixed temperature. On a masked graph, recall@10 exceeds degree-matched permutation (0.163 versus 0.072), but recovery is limited to reachable graph associations, not clinical prediction.

Summary by alphaXiv.

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