Substitution model validation consists of three hypothesis tests - a bootstrap test and two magional tests. The bootstrap test compares the observed and expected counts of site patterns including the sites with missing characters.
The first marginal test of base compositions between pairs of taxa can detect heterogeneity of base compositions across species, when the observed base frequencies of individual species deviate significantly from the overall average base frequencies across species.
The second marginal test is applied to the frequencies of double-nucleotides (doublets) between pairs of species to find pairs of species for which doublet frequencies are inconsistent with the pre-selected substitution model
To implement the three tests, users need to upload the alignment file(s) in phylip format. If the alignment file is in fasta or nexus format, please use "data conversion tool" to convert it to phylip format. Multiple alignment files are allowed but must be uploaded as a zip file and make sure all alignment files are in phylip format. After submission, a unique web link will be generated for users to download the result of the analysis.