Data-Driven Soft Labeling Scales DNA Read Classification to Whole-Body Cell-Type Deconvolution

Abstract:Revised following peer review. We expanded baseline comparisons, corrected evaluation leakage and read-boundary handling, clarified the confidence-weighted loss, and added sensitivity analyses for pooling and region selection. We also expanded TCS failure-mode and limitations analyses, added a discussion section, and provided code and data links for reproducibility.

Machine Learning (cs.LG); Genomics (q-bio.GN); Quantitative Methods (q-bio.QM)

arXiv:2607.04987 [cs.LG]

(or arXiv:2607.04987v3 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2607.04987 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.