Algorithm picks best mix of top ranking and diverse chemical compounds
OPDiv: Optimal Selection of Top-K High-Scoring, Diverse Compounds
Machine Learning
Summary
When searching for new chemicals, scientists find many good candidates but can only test a few. Picking the very best often means many similar compounds, while choosing diverse ones might miss the top scorers. The authors present OPDiv, an algorithm that finds the best balance by choosing a small group of chemicals that score high and are diverse. OPDiv uses a smart math approach called integer optimization to do this efficiently. This helps compare and improve virtual screening methods by focusing on diverse top picks, not just top scores.
What this means in practice
- •For pharmaceutical screening teams: Select compound sets balancing high predicted activity and diversity to improve drug discovery efficiency.
- •For chemical database curators: Create diverse but high-quality subsets from large chemical libraries for resource-efficient testing or analysis.
Authors
Miroslav Lžičař
Abstract
A virtual screening campaign may produce thousands of promising candidates, but only a small number can be purchased, synthesized, or tested. The practical question is how to select a set of compounds that both rank well and are diverse enough: this poses a genuine tradeoff, where selecting the highest-scoring molecules yields limited diversity, while diversity selection sacrifices some well-scoring molecules. We introduce OPDiv, a diversity selection and evaluation algorithm solving this tradeoff by finding an optimal subset of molecules using integer optimization. We demonstrate the selection algorithm in practice with fingerprint distance, shape and electrostatic diversity and compare the resulting diversity spectra. We argue that virtual screening is not merely a ranking problem, but also an implicit constrained optimization task: when redundant chemotypes are undesirable, pipelines should be compared based on the top-k compound selections satisfying the desired diversity constraints. OPDiv makes it possible to find the optimal compound set under a given diversity threshold efficiently and serves as a fair benchmark of the best diverse selection achievable by a given structure-based or ligand-based virtual screening pipeline, molecular search or generative model.