Optimal transport fate mapping resolves T cell differentiation dynamics across tissues

Immune responses evolve across time and tissues through coordinated programs of proliferation, differentiation, and migration, yet most single-cell measurements capture only static molecular snapshots. As a result, reconstructing how immune cells transition between alternative fates remains challenging, particularly for CD8 T cells, whose differentiation is highly dynamic and shaped by rapid expansion, contraction, and tissue trafficking. Here, we introduce an optimal transport-based fate mapping framework that reconstructs continuous CD8 T cell trajectories across time and tissues. Applied to longitudinal single-cell RNA-seq data from CD8 T cells responding to acute viral infection in mice, this approach accurately recapitulates population dynamics and resolves coherent effector and memory T cell differentiation trajectories. Extending the model to multiple tissues, we identify and experimentally validate temporally distinct waves of migration into the small intestine that give rise to divergent tissue-resident memory (TO_SCPLOWRMC_SCPLOW) fates, long-lived T cells crucial in immunosurveillance. By integrating optimal transport inference with time-resolved in vivo labeling, we demonstrate that CD52 marks recent tissue entrants and distinguishes them from TO_SCPLOWRMC_SCPLOW precursors. Finally, trajectory-guided analysis of transcription factor regulons reveals both shared and context-specific gene regulatory programs and identifies AP4 as a key regulator of circulating versus tissue-resident specification. These results establish optimal transport as a principled framework for reconstructing immune cell fate dynamics and provide a quantitative map of early events governing antiviral CD8 T cell differentiation across tissues.