22 A Model of Evolutionary Change in Proteins

In the eight years since we last examined the amino acid exchanges seen in closely related proteins,' the information has doubled in quantity and comes from a much wider variety of protein types. The matrices derived from these data that describe the amino acid replacement probabilities between two sequences at various evolutionary distances are more accurate and the scoring matrix that is derived is more sensitive i n detecting distant relationships than the one that we previously deri~ed.2,~ The method used 'in this chapter is essentially the same as that described in the Atlas, Volume 34 and Volume 5.' An accepted poinfmutation in a protein is a replacement of one amino acid by another, accepted by natural selection. I t is the result of two distinct processes: the first is the Occurrence of a mutation in the portion of the gene template producing one amino acid of a protein: the second is the acceptance. of the mutation by the species as the new predominant form. To be accepted, the new amino acid usually must function in a way similar to the old one: chemical and physical similarities are found between the amino acids that are observed to interchange frequently. Any complete discussion of the observed behavior of amino acids in the evolutionary process must consider the frequency of change of each amino acid to each other one and the propensity of each to remain unchanged. There are 20 X 20 = 400 possible comparisons. To collect a useful amount of information on these, a great many observations are necessary. The body of data used in this study includes 1,572 changes i n 71 groups of closely related proteins appearing in the Atlas volumes through Supplement 2. The mutation data were accumulated from the phylo-genetic trees and from a few pairs of related sequences. The sequences of all the nodal common ancestors in each tree are routinely generated. Consider, for example, the much simplified artificial phylogenic tree of Figure 78. 345 The matrix of accepted point mutations calculated from this tree is shown in Figure 79. We have assumed that the likelihood of amino acid X replacing Y is the same as that of Y replacing X, and hence 1 is entered in box YX as well as in box XY. This assumption is reasonable, because this likelihood should depend on the product of the frequencies of occurrence of the two …