fastTIGER: A rapid method for estimating evolutionary rates of sites from large datasets

Abstract

The evolutionary processes vary among sites of an alignment, called rate heterogeneity, that must be properly handled when analyzing the evolutionary relationships among species based on their genomic data. To this end, methods have been proposed to estimate the relative evolutionary rates between sites. Tree Independent Generation of Evolutionary Rates (TIGER) is a popular method to estimate the evolutionary rates among sites. However, the TIGER method is computationally expensive to calculate the evolutionary rates for large datasets, especially for whole genome datasets. In this paper, we present a simplified, fast, and accurate method, called fastTIGER, to estimate evolutionary rates for large datasets. Experiments on several large real datasets show that the evolutionary rates from the fastTIGER method have a reasonable correlation with ones estimated from the TIGER method while the fastTIGER method is several orders of magnitudes faster than the TIGER method. Moreover, the site rates estimated by fastTIGER method are as good as the ones estimated from the TIGER method in partitioning alignments to build maximum likelihood trees. The fastTIGER method enhances us to study the evolutionary relationships among species using their genomic data.