QSPR Models for the Prediction of Heats of Combustion in Organic Compounds Present in Foods

MARIO G. DIAZ; LUCAS A. GARRO; FRIDA C. D. DIMARCO PALENCIA; MATIAS F. ANDRADA; ESTEBAN G. VEGA HISSI; PABLO DUCHOWICZ; JUAN C. GARRO MARTINEZ

San Juan

Congreso; XLI Reunión Científica Anual de la Sociedad de Biología de Cuyo; 2023

Sociedad de Biología de Cuyo (SBC)

In the field of food research, the determination of heats of combustion (ΔcH) of nutrients is essential to estimate the amount of energy obtained for the metabolism during digestion. In this job, we performed a study of Quantitative Structure-Property Relationship (QSPR) models for the prediction of ΔcH values of three different compounds families, organic acids, amino acids and sulfur compounds, commonly found in various foods. For the development of the QSPR models, we used an experimental dataset from the bibliography of 130 compounds; 71 organic acids, 28 amino acids and 31 sulfur compounds. The molecular structures were characterized by 16000 molecular descriptors (MD) calculated by PaDEL software. The MD with best correlation with ΔcH were found by multiple lineal regression (MLR). The QSPR models demonstrated a high level of predictive capacity, with determination coefficients (R²) major to 0.8 and root mean square error (RMSE) values next to 0.1. In addition, the obtained mathematical equations have easy interpretation incorporating only two or three variables (MD). The results suggest that our QSPR models could be utilized for the prediction of the ΔcH of a wide range of compounds present in the foods and diets.