Bioptic B1: A Target-Agnostic Potency-Based Small Molecules Search Engine

TL;DR

This paper introduces Bioptic B1, a target-agnostic molecule search engine that efficiently screens ultra-large chemical libraries for biologically active, structurally diverse molecules using an optimized retrieval system.

Contribution

The paper presents a novel, efficacy-based, target-agnostic molecule search model combined with a fast, processor-optimized retrieval system capable of handling 40 billion molecules with high recall.

Findings

Achieved 100% recall rate on ultra-large chemical library

Outperformed state-of-the-art models in speed and retrieval quality

Enabled efficient virtual screening of billions of molecules

Abstract

Recent successes in virtual screening have been made possible by large models and extensive chemical libraries. However, combining these elements is challenging: the larger the model, the more expensive it is to run, making ultra-large libraries unfeasible. To address this, we developed a target-agnostic, efficacy-based molecule search model, which allows us to find structurally dissimilar molecules with similar biological activities. We used the best practices to design fast retrieval system, based on processor-optimized SIMD instructions, enabling us to screen the ultra-large 40B Enamine REAL library with 100\% recall rate. We extensively benchmarked our model and several state-of-the-art models for both speed performance and retrieval quality of novel molecules.