This Repository contains my final degree project (Degree in Biology) titled:

"TRATAMIENTO DE DATOS QUÍMICO-FORENSES PARA LA DISCRIMINACIÓN DE FLUIDOS BIOLÓGICOS EN MATERIALES SUPERABSORBENTES"

Autor: Ignacio Pachón Jiménez Tutor/es: Carmen García Ruiz

Año 2016

Abstract In this thesis, the objective pursued is to analyze the spectral data obtained by infrared spectroscopy from biological fluid stains on super-absorbent materials, in order to discriminate the presence of semen, which is a fluid of interest in many sexual crimes. The study consisted on the statistical analysis of the spectra obtained by infrared spectroscopy with Fourier transformed and attenuated total reflection (ATR-FTIR) of samples from several biological fluids (vaginal fluid, urine and semen), on a series of commercial super-absorbent materials (pads, salvaslips, and diapers). The spectra analysis has been performed by: baseline correction,normalization, and smoothing followed by chemometric techniques like Principal Component Analysis (PCA) and Pearson’s correlation coefficient. For the validation of the model, ROC curves were generated. The statistical procedure has been executed entirely in R software. The results show that the PCA does not allow the differentiation between semen and vaginal fluid. However, the applied Bayesian statistic yields a high efficiency for the discrimination of these fluids. Finally, the possibility of carrying out an enhanced methodology for the determination of the presence of semen on different super-absorbent materials, mixed with other biological fluids (vaginal fluid and urine), by infrared spectroscopy, has been confirmed.

Keywords: Forensic chemistry, infrared spectroscopy, biological fluids, semen, vaginal fluid, urine, super-absorbent materials, qualitative analysis, Principal Component Analysis (PCA), ROC.

CREDITS IN TEXT: pages 26-31