Analysis of amino-acid sequences by statistical technique

Abstract

This paper reports the findings obtained through the statistical analysis of the thermodynamic measurements collected on the partially mutated anti-lysozyme antibody HyHEL-10. The data contained amino-acid sequences of 35 types of mutants of HyHEL-10 and corresponding thermodynamic measurements such as entropy and enthalpy. We examined the contribution of each of the mutated sites on VH and VL chains of HyHEL-10 to the change of measurements by using the generalized linear model (GLM). Following results were obtained: (1) the sites VH32, 33, 50 and temperature had high partial correlations to the change of entropy (DH) and enthalpy (TDS), (2) the sites VH53 and 58 had high partial correlations to the change of specific heat (DCp), (3) the sites VL31, 32 and temperature had high partial correlations to the change of free energy (DG), DH, TDS and DCp. The results also suggested that the behavior of DH, TDS and DCp were well represented by GLM, however, the behavior of combining coefficient ( K a) should have a statistical nature assumed in Weibull distribution.