Data for: Transcript-wide association study identifies genes associated with aerobic and anaerobic intervention and prospective free-living active energy expenditure

Physical activity is associated with decreased risk for many chronic and acute conditions and is increasingly used as a therapeutic intervention for diabetes, cardiovascular disease among others. In this study, we aimed to identify common changes in expression of genes due to aerobic and anaerobic physical activity and investigated the efficacy of the identified exercise intervention signature in predicting prospective “real life” activity and sedentary behavior. We performed a multi-cohort analysis of twenty publicly available datasets (N=638) and identified a 23-gene set that discriminated between pre- and post- exercise in four independent interventional cohorts (N=245) validation sets with an average area under the curve (AUC) of 0.75. Our results suggest that our PA signature can be used be used as a means to assess physical activity intervention aggregated over a period of time, irrespective of the exercise modality and intensity of exercise.

The data deposited here represent the 638 samples that were used for Discovery of the signature and the 245 samples that were used for validation. All of these datasets are publicly available Gene Expression Omnibus. The dataset included here provides the original Gene Expression Omnibus accession number (GSE), together with high level details on the protocol, the total number of samples available, the exercise protocol, the sample type. We also include information on the number of samples that were used in our analysis to identify the 23 gene signature of physical activity together with the phase (Discovery vs Validation) information.