Learning Cross-Domain Representations for Transferable Drug Perturbations on Single-Cell Transcriptional Responses

TL;DR

This paper introduces XTransferCDR, a generative framework that learns transferable drug perturbation representations from single-cell transcriptomic data, improving phenotypic drug discovery across different domains.

Contribution

The paper presents a novel domain separation and cross-transfer approach for learning transferable drug perturbation features from transcriptomic data.

Findings

XTransferCDR outperforms existing methods on multiple datasets.

The model effectively decouples perturbation features from basal states.

Transferable representations enhance phenotypic drug discovery.

Abstract

Phenotypic drug discovery has attracted widespread attention because of its potential to identify bioactive molecules. Transcriptomic profiling provides a comprehensive reflection of phenotypic changes in cellular responses to external perturbations. In this paper, we propose XTransferCDR, a novel generative framework designed for feature decoupling and transferable representation learning across domains. Given a pair of perturbed expression profiles, our approach decouples the perturbation representations from basal states through domain separation encoders and then cross-transfers them in the latent space. The transferred representations are then used to reconstruct the corresponding perturbed expression profiles via a shared decoder. This cross-transfer constraint effectively promotes the learning of transferable drug perturbation representations. We conducted extensive evaluations of our model on multiple datasets, including single-cell transcriptional responses to drugs and single- and combinatorial genetic perturbations. The experimental results show that XTransferCDR achieved better performance than current state-of-the-art methods, showcasing its potential to advance phenotypic drug discovery.