TempoBridge: Source-Conditioned Flow Matching with Optimal Transport Couplings for Single-Cell Population Transitions

2026-10-04

AI Summary

This study addresses the challenges of unpaired data and insufficient source population discrimination arising from destructive single-cell measurements by proposing a unified source-conditioned transport framework. The method initializes latent transport with source cells and predicts population transitions via flow-time-generated velocity fields. By integrating source-conditioned flow matching with mini-batch optimal transport coupling, it achieves efficient training while eliminating the need for target expression or optimal transport computation during inference. Experimental results demonstrate that the proposed framework significantly outperforms existing baselines across diverse scenarios, including donor variation, gene perturbation, and compound treatment. These findings validate its strong cross-scenario generalization capability and practical utility for modeling single-cell dynamics.

Abstract

Destructive single-cell measurements provide unpaired population snapshots rather than observations of the same cells across conditions. Local cell states and transition requests may also be insufficient to distinguish responses across source populations. We introduce TempoBridge, a common source-conditioned transport formulation for temporal, genetic, and chemical population transitions. Source cells initialize latent transport and provide a fixed empirical population summary. The velocity field receives this summary alongside the evolving cell state, flow time, and a structured transition descriptor. Minibatch optimal transport (OT) supplies couplings only for conditional flow-matching training paths; inference requires neither target expression nor OT computation. On held-out donors, TempoBridge achieves an Energy distance of 0.129 versus 0.144 for scGen. Genetic mean-expression $L_2$ error is 2.261 versus 3.156 for scGPT-scratch under Seen 2/2. On held-out compounds, condition-averaged drug-effect correlation is 0.598 versus 0.561 for the CellFlow adapter. Temporal ablations show higher mean distributional error after removing source context, replacing optimal transport with random pairing, or replacing flow matching with static residual regression. Together, these results demonstrate the predictive utility of a common source-conditioned transport formulation across held-out donors, gene combinations, and compounds.

Problem

• single-cell population transitions
• unpaired population snapshots
• source-conditioned transport
• optimal transport couplings
• flow matching

Innovation

• Source-Conditioned Flow Matching
• Optimal Transport Couplings
• Single-Cell Population Transitions
• Conditional Velocity Field
• Minibatch Optimal Transport