647 Genome-wide Scans Identified Genetic Variants Associated with Facial Aging Traits Quantified by Deep Learning Methods

F. Wang,Q. Qi,Z. Li,X. Hu,R. Ye,L. Du,S. Wang
DOI: https://doi.org/10.1016/j.jid.2022.05.658
2022-01-01
Abstract:As life expectancy and aging population increase worldwide, facial aging is becoming progressively important. The occurrence and progression of facial aging are strongly influenced by genetic factors. Previous studies have identified a number of genetic variants associated with facial aging phenotypes using traditional image analysis and human assessments. In this study, we developed a deep learning framework to precisely measure 4 facial aging phenotypes (lacrimal sulcus, pigmentation spots, wrinkle forehead, and wrinkle under eyes). We then derived the phenotypes in 7,347 Han Chinese from two independent cohorts: Jidong cohort (N=5,036) and the National Survey of Physical Traits cohort (NSPT, N=2,311). A genome-wide scan in the Jidong cohort identified eight genomic regions showing genome-wide significant association with facial aging, including four previously unreported loci (i.e. 10q21.3, 9q21.12, 13q14.2 and 22q12.2) and four previously known genes: PPARGC1B, BNC2, MFSD12 and MC1R. All the eight loci were successfully replicated in the NSPT cohort at nominal significance. Meta-analysis of all samples further enhanced the significance level of the novel associations on 10q21.3, 9q21.12, and 22q12.2 (P=3.65×10-13, 2.82×10-8 and 5.00×10-9, respectively). In conclusion, our study identified novel genetic risk factors for facial aging traits using deep learning-based measurement in Han Chinese.
What problem does this paper attempt to address?