A rapid in vitro enzymic and chromatographic predictive model for the in vivo rat-based protein efficiency ratio of mixed food proteins.

A nutritional quality index in the nature of an enzymatic protein efficiency ratio (E-PER) was computed from the amino acid data of 18 different food products by means of multiple regression equations. The regression was performed by setting amino acid values derived from enzymic hydrolysis of the food proteins as the independent variables with the rat-based PER values as the dependent variables. The multiple regression gave the following equation: E-PER = -3.02 + 0.14 (asp) + 0.15 (glu) - 0.18 (pro) + 0.14 (ala) + 0.52 (met) + 0.21 (lys) + 0.09 (arg) - 0.45 (trp). The multiple correlation coefficient for this regression was 0.942 and the coefficient of variation was 88.7%. The prediction equation was tested on amino acid-PER data of 22 different foodstuffs and it successfully predicted (+/- 0.22) the PER of 17 and an effectiveness of 77.3%.