MapPFN: Learning Causal Perturbation Maps in Context

TL;DR

MapPFN is a meta-learning model that predicts causal perturbation effects in biological systems, adapting to new contexts using in-context learning and synthetic priors, with zero-shot and fine-tuned performance.

Contribution

It introduces MapPFN, a prior-data fitted network that leverages synthetic causal priors and in-context learning to adapt to unseen biological contexts at inference time.

Findings

Zero-shot performance matches models trained on real data.

Fine-tuning enhances prediction accuracy across contexts.

Code and data are publicly available at the provided URL.

Abstract

Planning effective interventions in biological systems requires treatment-effect models that adapt to unseen biological contexts by identifying their specific underlying mechanisms. Yet single-cell perturbation datasets span only a handful of biological contexts, and existing methods cannot leverage new interventional evidence at inference time to adapt beyond their training data. To meta-learn a perturbation effect estimator, we present MapPFN, a prior-data fitted network (PFN) pre-trained on a synthetic biological prior with causal interventions, decoupling pre-training from limited wet-lab data. Unlike existing methods, MapPFN uses in-context learning to map a sequence of experiments to a post-perturbation distribution, enabling a single pre-trained model to adapt to new datasets and arbitrary gene sets at inference time. Zero-shot, MapPFN identifies differentially expressed genes on par with models trained on real single-cell data, and fine-tuning further improves predictions across biological contexts. Our code, model and data are available at https://marvinsxtr.github.io/MapPFN.