Biobank Lab in cooperation with Department of Anthropology (both - University of Lodz) have developed new model for hair colour prediction in polish population. Obtained results were patented. The part of used results are presented in paper which is published in HOMO - Journal of Comparative Biology (IF=0,579). Enjoy your lecture.

 

"Association of five SNPs with human hair colour in the Polish population."

 

Anna Siewierska-Górska1, Aneta Sitek2, Elżbieta Żądzińska2, Grzegorz Bartosz1, Dominik Strapagiel3.
1Department of Molecular Biophysics, University of Lodz, Lodz, Poland.
2Department of Anthropology, University of Lodz, Lodz, Poland.
3Biobank Lab, Department of Molecular Biophysics, University of Lodz, Lodz, Poland.

 

 
 

Twenty-two variants (single nucleotide polymorphisms – SNPs) of the genes involved in hair pigmentation (OCA2, HERC2, MC1R, SLC24A5, SLC45A2, TPCN2, TYR, TYRP1) were genotyped in a group of 186 Polish participants, representing a range of hair colours (45 red, 64 blond, 77 dark). A genotype-phenotype association analysis was performed.

Using z-statistics we identified three variants highly associated with different hair colour categories (rs12913832:A > G in HERC2, rs1805007:T > C and rs1805008:C > T in MC1R). Two variants: rs1800401:C > T in OCA2 and rs16891982:C > G in SLC45A2 showed a high probability of a relation with hair colour, although that probability did not exceed the threshold of statistical significance after applying the Bonferroni correction. We created and validated mathematical logistic regression models in order to test the usefulness of the sets of polymorphisms for hair colour prediction in the Polish population. We subjected four models to stratified cross-validation. The first model consisted of three polymorphisms that proved to be important in the associative analysis. The second model included, apart from the mentioned polymorphisms, additionally rs16891982:C > G in SLC45A. The third model included, apart from the variants relevant in the associating analysis, rs1800401:C > T in OCA. The fourth model consisted of the set of polymorphisms from the first model supplemented with rs16891982:C > G in SLC45A and rs1800401:C > T in OCA. The validation of our models has shown that the inclusion of rs16891982:C > G in SLC45A and rs1800401:C > T in OCA increases the prediction of red hair in comparison with the algorithm including only rs12913832:A > G in HERC2, rs1805007:T > C and rs1805008:C > T in MC1R. The model consisting of all the five above-mentioned genetic variants has shown good prediction accuracies, expressed by the area under the curve (AUC) of the receiver operating characteristics: 0.84 for the red-haired, 0.82 for the dark-haired and 0.71 for the blond-haired.

A genotype-phenotype association analysis brought results similar to those in other studies and confirmed the role of rs16891982:C > G, rs12913832:A > G, rs1805007:T > C and rs1805008:C > T in hair colour determination in the Polish population. Our study demonstrated for the first time the possibility of a share of the rs1800401:C > T SNP in the OCA2 gene in hair colour determination. Including this single nucleotide polymorphism in the actual hair colour predicting models would improve their predictive accuracy.

 

This work was supported by the European Regional Development Fund (Project POIG 01.01.02-10-005/08).

KONTAKT
NAPISZ DO NAS

{mosmap height='270'|width='100%'|zoom='17'| lat='51.7767961'| lon='19.4843932'}