Environmental adaptation of proteins: regression models with simple physicochemical properties.

Bio-sequences from ortholog proteins are well suited for statistical inference when the sequences can be divided into ordinal groups based on known environmental features or traits of the host organisms. In this paper two new regression models are described for extracting proteomic trends of extreme environments. The approach is based on physicochemical properties of the amino acids, and may also utilise stratification of the data. We are especially looking for connections of temperature adaptation between the organism and its molecular level. To show the applicability of the methods, we present analyses of genomic data from proteobacteria orders, where we examine the cold adaptation of membrane proteins, intracellular proteins, and the enzyme endonuclease I. Our results confirm earlier findings that redistribution of charge and increase of surface hydrophobicity might be some of the most important signatures for cold adaptation.