Prediction of the antioxidant capacity of maize (Zea mays) hybrids using mass fingerprinting and data mining

Hybrid maize (Zea mays L.) is one of the major grains used as food for humans and animals. Besides carbohydrates and amino acids, maize contains micronutrients such as vitamins and phytochemicals with health-promoting effects. In a current maize breeding program, the phytochemical profiles of 35 maize hybrids grown in three significant agroecologies of Mexico highlands, subtropics and tropics were evaluated. For the different locations, phytochemical traits, which are relevant for nutraceutical effects, and the antioxidative capacity, were determined. Statistical analyses indicate the importance of the pedigree/agroecology on the phytochemical profiles. Mass fingerprinting in combination with the Random Forest Tree algorithm allowed the building of models that predicted the antioxidative capacity of the maize hybrids. The methods are, therefore, suitable for the efficient selection of maize hybrids with improved nutraceutical value.