TESTING THE CONSTANT-RATE NEUTRAL ALLELE MODEL WITH PROTEIN SEQUENCE DATA.

摘要: Langley and Fitch (1973, 1974) (1976) statistically analyzed the pattern of nucleotide substitutions in seven proteins 17 taxa rejected null hypothesis a constantrate Poisson model protein evolution. Their rejection constant-rate was based on extremely large observed value statistic (henceforth referred to as X2LF), which has standard Chi-squared distribution under their hypothesis. Gillespie (1979) showed that neutral allele model, X2LF is not distributed. Though they did obtain exact Of X2LF, were able show expectation an increasing function 0 = 4Nu, where N population size u mutation rate. Thus, analysis does constitute test model. Evidently no possible using only since matter how sufficiently could account for observation. If one knew 0, use estimate 0. I report here results Monte Carlo simulations used characterize several values The determine mean over many simulation trials equals X2LF. also Omil, 6 approximately 5% produced greater than or equal val-