BaCaDI: Bayesian Causal Discovery with Unknown Interventions
This addresses a key limitation in causal discovery for domains like biology where intervention targets are often unknown, enabling more reliable inference from limited experimental data.
The paper tackles the problem of inferring causal structures from data with uncertain or unknown intervention targets, such as in single-cell biology, by proposing BaCaDI, a Bayesian framework that jointly infers intervention targets and causal graphs. In experiments on synthetic and gene-expression data, BaCaDI outperforms related methods in identifying both causal structures and intervention targets.
Inferring causal structures from experimentation is a central task in many domains. For example, in biology, recent advances allow us to obtain single-cell expression data under multiple interventions such as drugs or gene knockouts. However, the targets of the interventions are often uncertain or unknown and the number of observations limited. As a result, standard causal discovery methods can no longer be reliably used. To fill this gap, we propose a Bayesian framework (BaCaDI) for discovering and reasoning about the causal structure that underlies data generated under various unknown experimental or interventional conditions. BaCaDI is fully differentiable, which allows us to infer the complex joint posterior over the intervention targets and the causal structure via efficient gradient-based variational inference. In experiments on synthetic causal discovery tasks and simulated gene-expression data, BaCaDI outperforms related methods in identifying causal structures and intervention targets.