We have read tens of thousands of publications and manually collected cell type signatures to create a unified, reliable, searchable resource. You can compare your cell type specific genes with individual marker gene sets from every reference publication instead of using an aggregated model for comparison.

Our data is continuously updated so you can use cell types from the latest publications to annotate cells in scRNA-seq datasets and perform spot deconvolution in spatial RNA-seq datasets. You don't have to rely on databases and models that have been updated months, or sometimes years, ago.

No need to search literature, download models, write code in R or Python, or upload your raw data. You can rapidly search through tens of thousands of cell type signatures from literature within seconds. Our rank-based method only needs a ranked gene list to accurately assign cell types to clusters from scRNA-seq and spatial RNA-seq datasets.

Cell type signatures from publications are tested for accuracy and reliability. We assign detailed experimental annotations related to canonical markers, tissues, diseases and experimental conditions to create an AI-ready resource.