Exploring the possibility of predicting human head hair greying from DNA using whole-exome and targeted NGS data

BMC Genomics

Table 5 Success rate in prediction of hair greying status in a total 849-sample set and in two extreme phenotypic groups; i) ≤30 years old and greying, ii) ≥40 years old and no greying

Correct predictions	Total N	Young (≤30 y.o.) and greying	Old (≥40 y.o.) and no greying
Age-based BNN model	677/848^a (79.8%)	0/71 (0.0%)	0/29 (0.0%)
12-variable BNN model	684/848^a (80.7%)	6/71 (8.5%)	0/29 (0.0%)
Age-based MNN model	609/822^b (74.1%)	0/68 (0.0%)	0/29 (0.0%)
14-variable MNN model	611/822^b (74.3%)	3/68 (4.4%)	2/29 (6.9%)

^aOne sample was discarded from BNN analyses due to the lack of information on age
^bOne sample was discarded from MNN analyses due to the lack of information on age and 26 additional samples were omitted due to the availability of information on binary status of hair greying only

ISSN: 1471-2164