Evaluating the discriminating power of amino acid ratios on distinguishing dark colored hair samples

Human hairs are one of the most commonly encountered items of trace evidence. Currently, conventional methods for hair analysis include microscopic comparison and DNA analysis (nuclear and mitochondrial). Each approach has its own drawbacks. Hair proteins are stable and offer an alternative to DNA testing, as demonstrated with proteomics for distinguishing humans. However, proteomics is complicated and requires identifying peptides to remain intact following harsh sample preparation methods. Alternatively, the actual amino acid content of a hair sample may also offer important identifying information and actually requires proteins and peptides to be broken down completely rather than remaining intact. This study evaluated the discriminating power of using hair amino acid ratios to differentiate hair samples from 10 unrelated individuals with dark colored hair. Hair proteins were digested, derivatized, and analyzed using gas chromatography-mass spectrometry. Amino acid ratios were calculated for each individual and comparisons using ANOVA and post-hoc pairwise t-test with Bonferroni correction were made with amino acid ratios for individuals. Overall, out of the 45 possible pairwise comparisons between all hair samples, 38 (84%) were differentiable. Out of the 36 possible pairwise comparisons between brown haired individuals, 32 (89%) were considered differentiable using univariate statistics. Multivariate statistics were also attempted but, overall, univariate models were sufficient for exclusionary purposes. These results indicate that amino acid ratio analysis can potentially be used as an exclusionary method using hair if DNA analysis cannot be performed, or to corroborate conclusions made following microscopic analysis.