Relational Anatomy of Cell Fate Across Lineage-Resolved Perturbation Experiments

Relational Anatomy of Cell Fate Across Lineage-Resolved Perturbation Experiments develops a finite relational calculus for distinguishing state–fate association, exact sufficiency, lineage fingerprinting, shared compression, repeated realization, persistent fate structure, and molecular realization across lineage-resolved single-cell experiments. The study analyzes public datasets spanning cancer resistance, hematopoietic differentiation, and cellular reprogramming, including ReSisTrace, Rewind, Watermelon, and CellTag-multi. Measured precursor states consistently refine future-outcome structure while retaining explicit mixed-fate fibers. Temporal history can reach exact finite recovery when the representation approaches lineage identity, and fate-blind compression tests whether that exactness survives in representations shared across distinct lineages. Repeated terminal realization yields a support-indexed fate hierarchy with reproducible whole-profile structure and heterogeneous terminal molecular inhabitation. A cross-carrier synthesis places 849 canonical state partitions in a common identity–future geometry. Exact recovery is concentrated at lineage-identifying representations: 352 exact surfaces are injective lineage fingerprints, while 18 are exact but noninjective, and none of the registered noninjective exact candidates survives the complete familywise control stack. At the same time, control-qualified fate-selective surplus shows that biological state can resolve future-discordant lineage pairs more strongly than generic lineage-identity resolution while residual future ambiguity remains. The resulting framework provides a finite, carrier-scoped language for separating predictive information, exact recovery, lineage identity, shared state–fate structure, replicate persistence, and terminal molecular realization. A public reproducibility repository accompanies the manuscript and contains the paper-specific derived artifacts, controls, validation scripts, figure-generation code, provenance records, and cross-carrier synthesis required to reproduce the reported empirical results.