Detecting heterogeneity of substitution along DNA and protein sequences.

摘要: Regions of differing constraint, mutation rate or recombination along a sequence DNA amino acids lead to nonuniform distribution polymorphism within species fixed differences between species. The power five tests reject the null hypothesis uniform is studied for four classes alternate hypothesis. explored are variance interval lengths; modified test, which includes covariance neighboring intervals; length longest interval; shortest third-order and composite test. Although there no uniformly most powerful test over range hypotheses tested, usually have highest power. Therefore, we recommend that one these two be used departure from uniformity in all circumstances. Tables critical values given. depend both on number events positions sequence. A computer program available request calculates specified as well significance level given data set.