Myofibroblastic CAFs arising from bone-resident osteoblast precursors retain an osteolineage signature and support breast cancer progression via Osterix-mediated signaling.

The codes are consolidated from RMarkdown files used for analyzing and visualizing data. Note: Publicly available data are retrieved from data availability section of the original publications. Patient metadata was retrived and organized into separate excel sheets from the supplementary data provided by the publication. The finalized codes for processing CAF subset from Wu 2021 dataset is no longer available due to untimely code backup and loss of cluster access. In principle, the integrated human CAF dataset can be reproduced by integrating the entire Wu 2021 and Cords 2023 datasets, removing low-quality cells, re-clustering, and isolating the PDGFRA+ PDGFRB+/- clusters. Clusters were annotated using canonical markers, which are shown in the publication but may not be included in the codes for preliminary processing of the datasets. Files required for performing ssGSEA over the TCGA datasets were generated from the bulk RNA and signature genesets manually and via command-line. The codes and procedurals are not included, but the files should be reproducible given the documentations on the file format.