Abstract With the rapid development of transcriptomics technology, gene expression profiles of normal human tissues have been frequently utilized as valuable references to study aging and disease mechanisms. Although individual databases have been established, most human tissues remain under-sampled. Meanwhile, profiles from RNA-seq and microarray platforms must be analyzed separately due to inherent distribution bias, further weakening the statistical power of health samples. Here, we constructed a comprehensive transcriptome atlas for human normal tissues (HuNT) containing 29,769 samples covering 74 human normal tissue types or parts, 70% and 37% larger than those in the well-known GTEx database in terms of sample size and tissue diversity, respectively. More importantly, instead of separate calculation, integrative analysis has been enabled for hybrid samples across RNA-seq and microarray profiles per tissue type. Based on the largest human sample collection so far, HuNT provides the latest information on: (1) tissue-specific genes, (2) stably and highly expressed housekeeping genes for each tissue, (3) gene expression landscape across all human tissues, and (4) top correlated genes for a query in a specified tissue. A case study of four widely-acknowledged biomarkers is also illustrated in the paper. With future data accumulation, HuNT may be useful to provide not only gene expression landscapes in human normal tissues, but also benchmarks to derive interesting marker genes and even therapeutic targets. HuNT is now freely accessible at http://hunt.badd-cao.net/ .
Tang et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: