Pipeline validation at the level of single residues. (A) TRIM5 analysis of 16 species with 96% coding sequence (CDS) coverage as compared with Sawyer et al. (2005) and van der Lee et al. (2017). (B)MAVS analysis of 19 species with 97% CDS coverage as compared with Patel et al. (2012). (C)SAMHD1 analysis of 18 species with 92% CDS coverage as compared with Laguette et al. (2012), Lim et al. (2012), and van der Lee et al. (2017). (D)MX1 analysis of 18 species with 97% CDS coverage as compared with Mitchell et al. (2012). See Table S3 for detailed analyses. Only the number of residues from M8 vs. M7 analysis are reported (see Materials and methods for details). Comparisons are made using similar phylogenetic branches of Simiiformes (hominoids, Old World monkeys, and New World monkeys) in most cases. Probability thresholds are set to match those used by the referenced studies. Number of species used varies due to the elimination of incomplete sequences from FREEDA analyses. Magenta ellipses, number of sites matching the threshold found by FREEDA; green ellipses, number of published sites. (E) Raw FREEDA-annotated image of the MxA protein. Note the region removed from the analysis due to alignment uncertainty (dark grey arrowhead) where four residues under positive selection have previously been identified (Mitchell et al., 2012).