Chemometric analysis of ATR-FTIR spectra of fingernail clippings for classification and prediction of sex in forensic context

Abstract

We report chemometric analysis of ATR-FTIR spectroscopy for classification and prediction of sex of male and female nail clippings from 100 individuals. An effort has been made for differentiation of fingernail clippings (50 each from males and females of age between 15 and 60 years). The spectral features through ATR-FTIR revealed the presence of prominent amide A & B band, amide I, amide II and amide III bands in the fingerprinting region. These bands confirmed the presence of keratin protein which is the main component present in fingernail clipping. PCA and PLS-DA were used for exploratory and predictive data analysis respectively. Before chemometric analysis, pre-treatment methods like normalization and obtaining derivative of spectra were performed on the raw data of the IR spectra. It was found that PCA could not differentiate the sexes well from these spectra, as such was only used for exploratory purposes. However, PLS-DA has been found to provide excellent classification and prediction with a high true positive rate for classification of sex. The correct classification rate from the normalised derivative data has been found to be 100% and 90% for females and males respectively, which is quite impressive from the forensic examination perspective. PLS-DA analysis after ATR-FTIR spectroscopy proves to be a suitable method for such type of forensic examination. This approach is a quick, non-invasive, and non-destructive method and can be used as an alternative technique to other conventional techniques.