Single-cell perturbation atlas of PTB prevention candidate drugs

Preterm birth (PTB) is the leading cause of mortality in children under five, yet effective preventive therapies are lacking. PTB’s multifactorial nature, as well as the high cost and ethical challenges of pregnancy clinical trials, hinder drug development efforts. Meanwhile, targeting immune pathways that regulate labor timing is emerging as a promising preventive strategy. We present Simulated Immune Modeling of Clinical Outcomes (SIMCO), a prediction ensemble that integrates single-cell immune perturbation modeling with outcome prediction to rapidly evaluate PTB prevention candidate drugs. To train SIMCO, we generated a mass cytometry perturbation atlas of 218 million leukocytes, capturing drug effects on specific cell types and signaling pathways that differed between pregnant and non-pregnant individuals. SIMCO accurately recapitulated the single-cell modulation induced by nine repurposed candidate drugs. When applied to an independent longitudinal pregnancy cohort, SIMCO simulated each drug’s effect on gestational length, offering a scalable framework for drug prioritization. Tetrahydrofolate, maprotiline, and the combination of aspirin and lansoprazole emerged as top candidates for PTB prevention, delaying labor onset in preterm samples by 9.2 (± 2.1), 9.2 (± 1.2), and 6.5 (± 1.8) days, respectively, primarily through enhanced mTOR signaling in innate immune cells and attenuated JAK/STAT signaling in naïve CD4⁺ T cells. This upload includes all raw .fcs files generated for the single-cell perturbation atlas pre-gated into 28 cell types via manual gating.