Answers are grounded in this paper and cite their sources.
Background Single-cell studies of colorectal cancer describe epithelial states and possible progression, but expression-state reproducibility, trajectory execution and progression-associated comparisons are often conflated. We evaluated these evidence layers using patients or donors, rather than cells, as the inferential units. Methods We applied a frozen patient-level framework to two prespecified epithelial expression programs. GSE294300 provided the independent paired normal-primary endpoint. GSE315534, GSE178318 and the Moorman 2024 cohort were analyzed separately for trajectory execution and descriptive progression-associated score changes. Complete-patient bootstrap intervals, coverage audits and retained failure/abstention records were used throughout. Results Sixteen patients with complete paired observations contributed to the GSE294300 endpoint. The mean paired difference for the absorptive program was − 0.20474 (patient-level bootstrap 95% confidence interval [CI], -0.26994 to -0.14409). The secretory-program mean paired difference was + 0.05494 (95% CI, -0.02083 to + 0.15804), with the interval crossing zero. Executable trajectory runs yielded finite estimates, whereas some method-resampling combinations abstained or failed. Progression-associated estimates differed across the three independent cohorts and were therefore reported separately rather than pooled. Conclusion The study provides an auditable framework for separating reproducible epithelial state summaries from trajectory execution behavior and cross-cohort progression generalization. The accompanying figures, tables, frozen contracts and reproducibility package support transparent patient-level analysis of heterogeneous colorectal cancer single-cell data.
No takes yet. Share an insight, caveat, or question.
Mu et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: