Fig. 2
From: Lung transcriptomic clock predicts premature aging in cigarette smoke-exposed mice
Development of age-prediction model from lung transcriptomics data (E-MTAB-5281, E-MTAB-5280 and E-MTAB-7444) using LASSO regression.. a. Random leave-out cross-validation over 100 runs. For every run, the 111 Sham samples from the 3 studies are randomly divided into training (75%) and validation sets (25%). b. Leave-one-study-out cross-validations. In each round, one study is excluded from the training but used for the validation. c. The final list of age predictor genes as selected by LASSO when applied to all of the 111 Sham transcriptomes. CS-regulated genes (see Methods) are colored red. MAE: mean average error