GEOapplication/xmlftp://ftp.ncbi.nlm.nih.gov/geo/series/GSE337nnn/GSE337237/primaryOK200GenomicsMus musculusGenome binding/occupancy profiling by high throughput sequencing Expression profiling by high throughput sequencinghttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE337237GEOGSEfalseSystematic Benchmarking of Ambient RNA decontamination Tools to Advance Precision in Single-Cell transcriptomic analysisAmbient RNA contamination in single-cell RNA sequencing introduces exogenous transcripts from lysed cells, distorting cell-type annotation and biological interpretation. We performed a systematic benchmark of 7 decontamination tools using datasets with well-defined ground truth. Our evaluation across accuracy, robustness, and subtype sensitivity reveals complementary strengths: CellBender and scCDC excel in contamination estimation accuracy, SoupX is the most robust and sensitive for rare subtypes. For overall decontamination efficacy, scAR best balanced robustness and accuracy at the cell-type level. Our study offers a practical guideline for context-dependent tool selection and highlights key directions for future algorithmic development.2026/08/10GSE337237GSM9851503GSM9851500GSM9851502GSM985150134290337237Mus musculus