DiSignAtlas

Molecular signatures can be used as diagnostic, prognostic, predictive, and therapeutic markers for a particular disease. They are usually sets of genetic variants, mRNAs, proteins, metabolites or other variables, and they can be derived from a broad range of omics data types. High-throughput sequencing technologies [e.g., microarray, RNA-Seq, single-cell RNA-Seq (scRNA-Seq) and single-nucleus RNA-Seq (snRNA-Seq)] enable quantitative evaluation of gene expression to identify signatures of differential gene expression in disease.

In DiSignAtlas, the raw data of 181,434 unique transcriptome profiles were manually collected and curated from well-conducted studies involving diseased and healthy humans and mice that are archived in GEO, ArrayExpress and TCGA. 10,306 datasets were organized related to 1,836 nonredundant disease types between the disease and control groups, including 328 scRNA-Seq/snRNA-Seq datasets. 5,498,991 differentially expressed genes were identified and deposited as signatures of human and mouse diseases in the DiSignAtlas.