Fig. 5
From: A population-based gene expression signature of molecular clock phase from a single epidermal sample
The pipeline for identifying and testing epidermal candidate biomarkers. The training dataset includes 483 longitudinal and population-based epidermis samples. Using this training dataset, a set of 12 candidate biomarker genes were selected by ZeitZeiger. The user-friendly SkinPhaser app was developed for predicting molecular clock phase using this set of 12 candidate biomarker genes. The prediction accuracy of candidate biomarkers was further tested using two testing datasets. One dataset includes 36 epidermis samples collected from 9 male participants in the USA, and each participant donated one sample at each of four time points (6 AM, 12 PM, 6 PM, and 12 AM). The other one includes 54 epidermis samples collected from 18 male and female participants in Germany, and each participant donated one epidermis sample at each of three time points (9:30 AM, 2:30 PM, and 7:30 PM)