Computational Tracking of Cell Origins Using CellSexID from Single-Cell Transcriptomes

These datasets support the validation and biological application of CellSexID, a computational method for inferring cell origin in sex-mismatched settings using single-cell RNA sequencing (scRNA-seq) data. CellSexID leverages the expression of sex-linked genes to classify individual cells as male or female, providing an in silico alternative to physical labeling. Two scRNA-seq datasets from sex-mismatched chimeric mouse diaphragm models are included: (i) a validation dataset, in which cell multiplexing labels were applied to establish ground-truth cell origin, and (ii) an experimental dataset, analyzed without labels to demonstrate the method's applicability in unlabeled settings. Together, these datasets provide a resource for benchmarking computational cell origin tracking methods and for studying immune cell populations in chimeric model systems. Overall design: The chimeric mouse diaphragm validation dataset for CellSexID was generated using sex-mismatched bone marrow chimeras. Female CD45.2 mice underwent whole-body irradiation with diaphragm shielding, followed by transplantation of male CD45.1 bone marrow. Diaphragm macrophages were isolated by FACS, labeled with Cell Multiplexing Oligonucleotides (female: CMO306; male: CMO305), and processed for single-cell RNA sequencing to provide a reference for cell origin classification. The chimeric mouse diaphragm experimental dataset was prepared using the same protocol, except macrophages were not sex-sorted; instead, cells from each mouse were pooled and labeled with a unique Cell Multiplexing Oligonucleotide to enable mouse-specific identification. In the experimental datasets, FACS sorting and cell preparation were performed on four mice per day over two consecutive days (Day 1 and Day 2), yielding a total of eight mice. All libraries were subsequently prepared and sequenced in a single batch.