Cross-modality Matching and Prediction of Perturbation Responses with Labeled Gromov-Wasserstein Optimal Transport

TL;DR

This paper introduces an advanced optimal transport method that aligns and predicts cellular responses across different measurement modalities, enabling integrated analysis of large-scale perturbation data in cell biology.

Contribution

It extends Gromov-Wasserstein Optimal Transport to include perturbation labels, facilitating cross-modality alignment and prediction of cellular responses, a novel approach in this context.

Findings

Effective cross-modality alignment demonstrated

Accurate prediction of cellular responses across modalities

Potential for unified causal models in cell biology

Abstract

It is now possible to conduct large scale perturbation screens with complex readout modalities, such as different molecular profiles or high content cell images. While these open the way for systematic dissection of causal cell circuits, integrated such data across screens to maximize our ability to predict circuits poses substantial computational challenges, which have not been addressed. Here, we extend two Gromov-Wasserstein Optimal Transport methods to incorporate the perturbation label for cross-modality alignment. The obtained alignment is then employed to train a predictive model that estimates cellular responses to perturbations observed with only one measurement modality. We validate our method for the tasks of cross-modality alignment and cross-modality prediction in a recent multi-modal single-cell perturbation dataset. Our approach opens the way to unified causal models of cell biology.