Peer Review #2 of "Potential use of human hair shaft keratin peptide signatures to distinguish gender and ethnicity (v0.1)"

Methods. Extraction of the human hair shaft proteins was performed using a newly developed alkaline solubilisation method. The extracts were profiled by 2-dimensional electrophoresis and resolved protein spots were identified by mass spectrometry and queried against the human hair database. The study was then followed-up by immunoblotting of the identified hair shaft keratin of interest using commercially available antibodies.

137 Data analysis 138 Silver-stained 2-dimensional electrophoresis gels were scanned using ImageScanner III (GE 139 Healthcare, Uppsala, Sweden) and analysis of protein spot volume was performed using 140 ImageMaster Platinum 7.0 software (GE Healthcare, Uppsala, Sweden). Image analysis was 141 restricted to protein spot clusters that appeared consistently within each group of hair shaft 142 proteins. The levels of proteins in each sample were calculated as a percentage of volume 143 contribution (% vol) in which the volume of contribution refers to the volume percentage of a 144 protein taken against the total spot volume of all the proteins, in order to eliminate the possible 145 variations due to staining and/or protein loading. Automatic spot detection was performed with 146 default parameters setting (cut-off parameters were Smooth-2; Saliency-147 1; Min area-5), whereas spot editing and removal of artefacts were done manually.

148
Data were analysed using the Statistical Package for Social Sciences (SPSS) version 25.0 149 (IBM Corporation, New York, USA). Shapiro-Wilk test of normality was used to assess the 150 distribution of the datasets. Parametric Student's t-test and ANOVA was used to analyse and 151 identify significant changes in expression between two population means (gender) and several 152 population groups (ethnicities), respectively, when the normality assumption was met (p > 0.05).
153 Following ANOVA, a pairwise comparison procedure was performed to test all possible 154 pairwise differences of the means using either Tukey HSD when equal variances was assumed 155 or, Dunnett's T3, if otherwise. On the other hand, non-parametric Mann-Whitney U and 156 Kruskal-Wallis tests with rank-based post-hoc test was used as their respective counterparts 270 earlier reported (Barthelemy et al., 2012;Thibaut et al., 2009;Nakamura et al., 2002). 271 Identification of these proteins by mass spectrometry and search of the human hair protein database 272 showed that they comprised the different types of keratins, the cysteine-rich helicoidal proteins 273 that protect the hair because of their insolubility and impermeability. However, all these spots of altered abundance appeared to be 281 truncated or processed keratin products as they were resolved within the molecular weight regions 282 lower than their putative primary translated precursor polypeptides (Table 1). The influence or 283 effects of processing of proteins in this material remains to be elucidated in future investigations. The gel electrophoresis profiling of 30 hair shaft samples taken from women of similar age 292 range but from three distinctive Malaysian ethnic subpopulations further showed significant 293 altered abundance of one type-I (K33b) and four type-II (K81, K83 and two K86) keratins between 294 the ethnic groups that were analysed. Like the earlier detected keratins, the type-I K33b and all 295 the type-II keratins detected are also known to be localised exclusively in the hair shaft (Moll, Divo   296 & Langbein, 2008). Based on the resolved experimental molecular weights, all the five spots of 297 altered abundance also appeared to be truncated or processed keratins. In addition, the K81 and 298 K86 spots that were also detected in this analysis were different from their counterparts that were 299 detected in the earlier gender analysis as they showed distinctive experimental molecular weight 300 and pI values.

301
The results of the latter study also demonstrated that the Indian ethnic group to be most 302 distinctive as they showed four abundantly different keratins (K33B, K81, K83 and K86 (spot 7)) 303 compared to the Chinese and three keratins that were differently altered (K33B and two K86 (spots      (Fig. 1).
Gel images were analysed by ImageMaster 2D Platinum Software (mean ± 95% confidence interval; n = 20). Panels A -F demonstrate the 6 protein spots (K85, three protein species of K86 and two protein species of K81) that were significantly different in abundance between male and female subjects. FC is fold change between the mean values for males and females. Error bars represent 95% confidence intervals.     Spots that were significantly different in abundance between subjects of distinct genders (spots 1-6) and ethnicities (spots 7-11) were excised from gels and subjected to in-gel trypsin digestion, MALDI-ToF/ToF analysis and human hair database query (Fig 1.). Experimental mass was calculated based on relative mobilities (R f ) of the spots. cov -coverage; pIisoelectric point.