In recent years, there has been an exponential increase in single-cell RNA sequencing (scRNA-seq) data and analysis methods. This advancement provides an opportunity for single-cell eQTL mapping, enabling the exploration of transcriptome heterogeneity across cell types at a more refined resolution (2018-Nat Genet, PMID: 29610479; 2022-Nature, PMID: 35545678).

However, to the best of our knowledge, there is currently no comprehensive database available to accommodate the growing number of single-cell-level TWAS results. This gap can be attributed to limited computing resources and the scarcity of eQTL models at single-cell level. Here we introduce the scTWASdb server to meet this need.

  • Construction of over 200 single-cell level eQTL models. We have systemically collected single-cell level eQTL datasets from publicly available sources, including individual- and summary-level reference eQTL datasets. Utilizing the OTTERS and TIGER methods (2023-Nat Comm, PMID: 36882394; 2019-Am J Hum Genet, PMID: 31230719), we have built over 200 models to date. These models cover 61 parenchymal cells and 143 immune cells, corresponding to 24 major cell types and 7 experimental treatments. Notably, the immune cells encompass various cell types across fluid functional cell states, namely dynamic eQTLs, which can provide valuable insights into gene regulation under specific simulation conditions.
  • Comprehensive GWAS summary data. We have collected extensive GWAS summary data from 18 different types of diseases, including 829 binary traits sourced from the UK Biobank. These traits were categorized according to the unique ICD-10 codes, a standardized multi-categorical system for diseases. We have conducted TWAS using each summary-level GWAS data with 204 imputation models trained from single-cell eQTL datasets. Both FUSION and S-PrediXcan test statistics have been included in the scTWASdb.
  • User-friendly web interface. We have developed a user-friendly web server ( that allows users to easily browse, search and download association information, relevant research metadata, and annotation information of interest. Additionally, an online tool for scTWAS analysis has been provided, enabling users to perform custom analyses using newly submitted GWAS summary data with our single-cell level models.
  • Free availability and local use of GWAS, single-cell eQTL, and scTWAS data. All data within scTWASdb are freely available and can be downloaded for local use.