Generative Intervention Models for Causal Perturbation Modeling

TL;DR

This paper introduces a generative intervention model (GIM) that predicts the effects of unseen perturbations in causal systems by mapping perturbation features to intervention distributions, improving mechanistic insights and out-of-distribution predictions.

Contribution

The paper presents a novel GIM approach that jointly learns causal mechanisms and predicts perturbation effects from features, outperforming existing causal inference methods.

Findings

GIM achieves robust out-of-distribution predictions on synthetic and biological data.

GIM effectively infers underlying perturbation mechanisms.

GIM performs on par or better than unstructured approaches in predictive accuracy.

Abstract

We consider the problem of predicting perturbation effects via causal models. In many applications, it is a priori unknown which mechanisms of a system are modified by an external perturbation, even though the features of the perturbation are available. For example, in genomics, some properties of a drug may be known, but not their causal effects on the regulatory pathways of cells. We propose a generative intervention model (GIM) that learns to map these perturbation features to distributions over atomic interventions in a jointly-estimated causal model. Contrary to prior approaches, this enables us to predict the distribution shifts of unseen perturbation features while gaining insights about their mechanistic effects in the underlying data-generating process. On synthetic data and scRNA-seq drug perturbation data, GIMs achieve robust out-of-distribution predictions on par with unstructured approaches, while effectively inferring the underlying perturbation mechanisms, often better than other causal inference methods.