Cremoso cheese: prediction of ripening time using physicochemical parameters and multivariate statistical techniques

Abstract

Physicochemical parameters were used to study the ripening evolution in Cremoso cheese. Twenty-eight cheeses manufactured in different seasons and ripened for 30 days were obtained from two nearby factories. Using principal components analysis (PCA), cheese samples were grouped by ripening time, mainly according to PC1. The high loading values of %pH4.6-SN/TN, %TCA-SN/TN, %PTA-SN/TN and the aS1-I-casein/(aS1-casein + aS1-I-casein) ratio on factor 1 indicate that these variables are the most important for grouping cheeses according to ripening time. Using Partial Least Squares Regression (PLS) based on physicochemical parameters selected by PCA, a good model for predicting the ripening time of Cremoso cheese was developed. This model includes parameters of moderate, simple, and fast analyses and has an error of estimation of 2 days.