Empowering Chemical Structures with Biological Insights for Scalable Phenotypic Virtual Screening

TL;DR

DECODE is a framework that enhances chemical structure representations with biological insights, enabling scalable and biologically meaningful virtual screening without requiring biological data at inference.

Contribution

This work introduces DECODE, a novel method that incorporates biological signals into chemical representations using limited paired data, improving drug similarity and discovery performance.

Findings

Over 20% improvement in mechanism-of-action prediction in zero-shot settings.

Six-fold increase in hit rates for new anti-cancer agents.

Effective filtering of experimental noise in biological data.

Abstract

Motivation: The scalable identification of bioactive compounds is essential for contemporary drug discovery. This process faces a key trade-off: structural screening offers scalability but lacks biological context, whereas high-content phenotypic profiling provides deep biological insights but is resource-intensive. The primary challenge is to extract robust biological signals from noisy data and encode them into representations that do not require biological data at inference. Results: This study presents DECODE (DEcomposing Cellular Observations of Drug Effects), a framework that bridges this gap by empowering chemical representations with intrinsic biological semantics to enable structure-based in silico biological profiling. DECODE leverages limited paired transcriptomic and morphological data as supervisory signals during training, enabling the extraction of a measurement-invariant biological fingerprint from chemical structures and explicit filtering of experimental noise. Our evaluations demonstrate that DECODE retrieves functionally similar drugs in zero-shot settings with over 20% relative improvement over chemical baselines in mechanism-of-action (MOA) prediction. Furthermore, the framework achieves a 6-fold increase in hit rates for novel anti-cancer agents during external validation. Availability and implementation: The codes and datasets of DECODE are available at https://github.com/lian-xiao/DECODE.