PerturbPFN: Probing the Limits of Synthetic Priors in Perturbation Modelling

Predicting cellular responses to unseen chemical perturbations is challenging due to unknown mechanisms, high-dimensional responses, and limited experimental coverage. We propose PerturbPFN, a PFN-style amortized model for unknown-target perturbation prediction under a hierarchical synthetic structural prior. Rather than directly regressing expression responses, PerturbPFN infers a latent system graph, sparse intervention targets, and intervention strengths, then propagates their effects through an SCM decoder. The model is trained entirely on prior-predictive synthetic episodes generated from biologically motivated graph and expression simulators, enabling structured in-context inference without test-time gradient updates. We evaluate PerturbPFN on both real single-cell perturbation data and synthetic benchmarks, covering effect prediction, target identification, and regulatory structure discovery. Our results show that PerturbPFN offers a complementary trade-off to specialized baselines, achieving competitive perturbation prediction with low inference cost while exposing interpretable intermediate estimates of targets, strengths, and system structure.