Identification of branched-chain amino acids in quinoa using linear regression modelling with machine learning.

Abstract

Quinoa is classified as a pseudocereal and contains bioactive components, such as proteins with high nutritional value. The consumption of protein and especially branched-chain essential amino acids plays a fundamental role in the diet, since it favors the maintenance of body structures. The present article seeks to perform an analysis and quantification of amino acids in quinoa protein using the United States Department of Agriculture (USDA) database. Python programming language was used to develop machine learning (ML) models to predict the presence of branched amino acids in quinoa protein and through linear regression to obtain statistical data that provide an approximate amount of these nutrients. Important articles can be found where information on machine learning techniques to obtain nutritional information is presented. The amino acids valine, leucine and isoleucine were identified in isolated quinoa protein and, in turn, chickpea, rice, broccoli, raw quinoa and cooked quinoa were used as controls, and the w/w concentration values obtained in g/100g were 15.9, 3.43, 1.7, 0.37, 2.1 and 0.66, respectively. These results show the presence of branched-chain amino acids in quinoa, confirming the importance of its nutritional value and its possible effects on muscle health.