cellPred is an easy-to-use interactive Shiny web app to help users with the identification of the underlying tissue or - at even higher resolution - actionable cell-types for a number of diseases. Moreover, one could additionally predict disease/phenotypic traits enriched in cell clusters that are obtained through single-cell RNA sequencing experiments. GWA studies have identified thousands of associations between genetic variants and various human disease traits. However, mapping such associations with disease mechanisms has been difficult as GWAS results are tissue/cell-type agnostic. Integration of single-cell RNA-seq cell-type signature may help identify the tissue/cell-type of action.
The cellPred is a Kolmogorov-Smirnov test based method which computes a mean specificity metric and uses a non-parametric test to evaluate enrichments between GWAS results and cell-type signatures.