Target-Free Compound Activity Prediction via Few-Shot Learning

TL;DR

This paper introduces a novel neural architecture for few-shot compound activity prediction that estimates continuous activity levels, outperforming existing methods in target-free drug discovery scenarios.

Contribution

The authors propose a new meta-learning model that predicts continuous compound activities using a specialized neural architecture, advancing beyond binary classification in few-shot settings.

Findings

FS-CAP surpasses traditional similarity-based methods.

The model effectively captures assay information.

It performs well across diverse datasets.

Abstract

Predicting the activities of compounds against protein-based or phenotypic assays using only a few known compounds and their activities is a common task in target-free drug discovery. Existing few-shot learning approaches are limited to predicting binary labels (active/inactive). However, in real-world drug discovery, degrees of compound activity are highly relevant. We study Few-Shot Compound Activity Prediction (FS-CAP) and design a novel neural architecture to meta-learn continuous compound activities across large bioactivity datasets. Our model aggregates encodings generated from the known compounds and their activities to capture assay information. We also introduce a separate encoder for the unknown compound. We show that FS-CAP surpasses traditional similarity-based techniques as well as other state of the art few-shot learning methods on a variety of target-free drug discovery settings and datasets.