scCT-DB, an omnibus for cancer patient-derived paired pre- and post-treatment single-cell transcriptomes to reveal drug perturbation and drug resistance mechanisms.

Drug resistance continues to be a major challenge in cancer treatment. Understanding cellular and molecular dynamics after treatment is crucial for elucidating resistance mechanisms. Single-cell RNA sequencing (scRNA-seq) of paired pre- and post-treatment patient samples enables high-resolution exploration of such dynamics, but rapidly accumulated relevant data bring challenges for easy data access, integration, and reuse. Therefore, we present the Cancer Treatment-related Single-Cell transcriptome DataBase (scCT-DB, http://scctdb.ncpsb.org.cn). scCT-DB has comprehensively collected 266 patient-derived paired pre- and post-treatment scRNA-seq datasets processed by a uniform pipeline and with detailed and structured metadata, most of which simultaneously include information on primary/acquired drug response. scCT-DB includes 6.19 million cells from 1142 original patient samples, 27 major cancer types, 96 therapeutic regimens, and 102 drugs (involving chemotherapy, immunotherapy, hormone therapy, and targeted therapy). scCT-DB also provides 48 dataset pairs with opposite primary response groupings and 41 longitudinal datasets with ≥3 sampling timepoints. Besides data browsing, download, and search, scCT-DB also supports single-dataset analysis (including cell abundance, gene, cell state, and intercellular communication perturbation analyses), dataset comparative and re-combination analyses, providing insights into drug perturbation mechanisms and their heterogeneity in patients, drug resistance mechanisms, and discovery of biomarkers predictive of treatment response, etc.