Code and processed data for AVF transcriptomic, machine-learning, and clinical analyses

The study integrates public transcriptomic datasets, WGCNA, machine-learning-based hub gene prioritization, single-cell RNA-sequencing analysis, immune infiltration analysis, CellChat analysis, and prospective clinical cohort validation to investigate the ALPL–ALP axis in AVF maturation failure. The deposited files include processed expression matrices, R analysis scripts, and de-identified clinical cohort data. The revised clinical dataset was curated by updating overlapping variables according to the verified source table while preserving newly added clinical variables related to liver disease, CKD-MBD, phosphate binder use, active vitamin D use, medication exposure, and liver function. These materials are intended to facilitate transparency and reproducibility of the analyses reported in the manuscript. Public datasets used in this study are available from GEO under accession numbers GSE220796, GSE119296, and GSE250469.