Causal Imputation via Synthetic Interventions

TL;DR

This paper introduces a causal imputation method using synthetic interventions to predict effects across many action-context pairs with limited experiments, validated on biological data.

Contribution

It extends synthetic interventions to handle data sparsity in causal imputation, providing theoretical guarantees under a latent factor model.

Findings

Outperforms standard baselines on the CMAP dataset

Provides valid causal estimates under a latent factor model

Addresses scalable prediction of compound effects across cell types

Abstract

Consider the problem of determining the effect of a compound on a specific cell type. To answer this question, researchers traditionally need to run an experiment applying the drug of interest to that cell type. This approach is not scalable: given a large number of different actions (compounds) and a large number of different contexts (cell types), it is infeasible to run an experiment for every action-context pair. In such cases, one would ideally like to predict the outcome for every pair while only having to perform experiments on a small subset of pairs. This task, which we label "causal imputation", is a generalization of the causal transportability problem. To address this challenge, we extend the recently introduced synthetic interventions (SI) estimator to handle more general data sparsity patterns. We prove that, under a latent factor model, our estimator provides valid estimates for the causal imputation task. We motivate this model by establishing a connection to the linear structural causal model literature. Finally, we consider the prominent CMAP dataset in predicting the effects of compounds on gene expression across cell types. We find that our estimator outperforms standard baselines, thus confirming its utility in biological applications.