Overlapping association signals in the genetics of hair-related phenotypes in humans and their relevance to predictive DNA analysis. (July 2022)

• Record Type: Journal Article
• Authors: Pośpiech, Ewelina; Karłowska-Pik, Joanna; Kukla-Bartoszek, Magdalena; Woźniak, Anna; Boroń, Michał; Zubańska, Magdalena; Jarosz, Agata; Bronikowska, Agnieszka; Grzybowski, Tomasz; Płoski, Rafał; Spólnicka, Magdalena; Branicki, Wojciech
• Is Part Of: Forensic science international. Volume 59(2022)
• Publication Date: 2022-07
• Subjects: Hair thickness -- Hairiness level of the body -- Monobrow -- Overlapping association signals -- Prediction modelling -- Forensic DNA Phenotyping; Forensic genetics -- Periodicals; Génétique légale -- Périodiques; Forensic genetics; Electronic journals; Periodicals; 614.1
• DOI: 10.1016/j.fsigen.2022.102693
• Languages: English
• ISSNs: 1872-4973

Abstract

Abstract: Genetic prediction of different hair phenotypes can help reconstruct the physical appearance of an individual whose biological sample is analyzed in criminal and identification cases. Up to date, forensic prediction models for hair colour, hair shape, hair loss and hair greying have been developed, but studies investigating predictability of hair thickness and density traits are missing. First data suggesting overlapping associations in various hair features have emerged in recent years, suggesting partially common genetic basis and molecular mechanisms, and this knowledge can be used for predictive purposes. Here we aim to broaden our understanding of the genetics underlying head, facial and body hair thickness and density traits and examine the association for a set of literature SNPs. We characterize the overlap in SNP association for various hair phenotypes, the extent of genetic interactions and the potential for genetic prediction. The study involved 999 samples from Poland, genotyped for 240 SNPs with targeted next-generation sequencing. Logistic regression methods were applied for association and prediction analyses while entropy-based approach was used for interaction testing. As a result, we refined known associations for monobrow and hairiness ( PAX3, 5q13.2, TBX ) and identified two novel association signals in IGFBP5 and VDR . Both genes were among top significant loci, showed broad association with different hair-related traits and were implicated in multiple interaction effects. Overall, for 14.7% of SNPs previously associated with head hair loss and/or hair shape, a positive signal of association was revealed with at least one hair feature studied in the current research. Overlap in association with at least two hair-related traits was demonstrated for 24 distinct loci. We showed that the associated SNPs explain ∼5–30% of the variation observed in particular hair traits and allow moderate accuracy of prediction. The highest accuracy was achieved for hairiness level prediction in females (AUC = 0.69 for the "none", 0.69 for the "low" and 0.76 for the "excessive" hairiness category) and monobrow (AUC = 0.69 for the "none", 0.62 for the "slight" and 0.70 for the "significant" monobrow category) with 33% of the variation in hairiness level in females explained by 7 SNPs and age, and 20% of the variation in monobrow captured by 7 SNPs and sex. Our study presents clear evidence of pleiotropy and epistasis in the genetics of hair traits. The acquired knowledge may have practical application in forensics, as well as in the cosmetic industry and anthropological research. Highlights: Pleiotropy and epistasis increase the complexity of human traits determination. Overlapping associations for hair-related traits demonstrated for 24 literature loci. Novel associations of the IGFBP5 and VDR genes with hair traits were discovered. Analysed SNPs explain ∼5–30% of the variation observed in particular hair traits. Hairiness in females and monobrow predicted with AUC= 0.69 and 0.68, respectively.