Prediction of Adipose Browning Capacity by Systematic Integration of Transcriptional Profiles

Activation and recruitment of thermogenic cells in human white adipose tissues (“browning”) can counteract obesity and associated metabolic disorders. However, quantifying the effects of therapeutic interventions on browning remains enigmatic. Here, we devise a computational tool, named ProFAT (profiling of fat tissue types), for quantifying the thermogenic potential of heterogeneous fat biopsies based on prediction of white and brown adipocyte content from raw gene expression datasets. ProFAT systematically integrates 103 mouse-fat-derived transcriptomes to identify unbiased and robust gene signatures of brown and white adipocytes. We validate ProFAT on 80 mouse and 97 human transcriptional profiles from 14 independent studies and correctly predict browning capacity upon various physiological and pharmacological stimuli. Our study represents the most exhaustive comparative analysis of public data on adipose biology toward quantification of browning after personalized medical intervention. ProFAT is freely available and should become increasingly powerful with the growing wealth of transcriptomics data. Cheng et al. develop a computational tool named ProFAT for quantifying the thermogenic potential of mouse and human fat samples based on automated and unbiased prediction of white and brown adipocyte content from raw gene expression datasets. ProFAT is freely available and can be accessed at http://profat.genzentrum.lmu.de.