To measure the evolutionary constraint on X-chromosomal genes for comparison between genes associated with XLMR, genes associated with diseases other than mental retardation, and genes not associated with disease, Ensembl macaque orthologs were identified and, where necessary, realigned using ClustalW2. We used DnaSP37 to determine the total and aligned length of each alignment, the number of synonymous positions and differences, and the number of nonsynonymous positions and differences. Genes for which a macaque ortholog was not found, or which showed an aligned length less than 90% of the total length, were omitted. From the remaining genes (72 genes associated with XLMR, 48 genes associated with other diseases and 393 genes not associated with disease, respectively), the ratio dN/dS (number of nonsynonymous changes per nonsynonymous site/number of synonymous changes per synonymous site) was calculated both for each individual gene and for all genes within each of the three categories concatenated (Supplementary Table 8 online). The significance of differences between the categories was evaluated using a Mann-Whitney test. To investigate whether any genes showed more missense variants within humans than expected from their evolutionary history, we used a McDonald-Kreitman test24.
