Reproduce “Automatic design of novel potential 3CL$^{\text{pro}}$ and PL$^{\text{pro}}$ inhibitors” 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
Timothy Atkinson, Saeed Saremi, Faustino Gomez, Jonathan Masci
MONAS searches for potential protease inhibitors by combining a graph neural network’s activity predictions with a learned energy score that favors molecules resembling its training data, then guiding molecule generation with Monte Carlo tree search. Using SARS-CoV-1 assay data, the system explored 40 million molecules and retained 120,000 predicted candidates; none were experimentally validated.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.