Reproduce “Explaining Chemical Toxicity using Missing Features” 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
Kar Wai Lim, Bhanushee Sharma, Payel Das, Vijil Chenthamarakshan, Jonathan S. Dordick
The authors apply contrastive explanations method (CEM) to molecular fingerprints, identifying present features that support a prediction and missing features whose addition changes it. For triphenylborane, five missing fingerprint bits, including bromine and hydroxyl, shift the model’s SR-MMP toxic-class probability from 0.0205 to 0.5062. These are classifier counterfactuals, not validated chemical edits, and concern only SR-MMP.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.