The oldest macaque fossil found in Asia is dated at approximately 5.5 million years old, not long after the presumed divergence of Asian from African macaques, suggesting that the migration to Asia was relatively rapid. However, our data do not allow us to pinpoint a location for the evolution of TRIMCyp. Although our data are most consistent with an origin of TRIMCyp in the common ancestor of Asian macaques, we have also considered several alternative hypotheses. First, it is possible that TRIMCyp was present in ancestral Old World primates but has been lost in all lineages other than Asian macaques. Results shown in Figure 4 clearly show that TRIMCyp sequences have not been lost at the DNA level, by deletion of the CypA sequence or by reversion of the exon 7 splice site. If this had occurred in some species, we would expect them to have the G allele but to group with the T-containing sequences, or to have the T allele in the absence of the CypA insertion. Neither of these features is present in any of the species (+)-JQ1 tested. Instead, our data show unambiguously that sequences containing the T allele form a monophyletic group, distinct from those containing the G allele. Thus, it is unlikely that TRIMCyp was lost at the DNA level in any lineage. In contrast, we cannot formally rule out the possibility that TRIMCyp was lost by lineage sorting. In our phylogenetic analysis, the T alleles appear to branch off before the separation of baboon and macaque G alleles. This could be taken to suggest that the T allele evolved before this evolutionary branching, and thus that TRIMCyp, or at least TRIMCyp-linked sequence changes, may be older than suggested by our other data. However, this analysis is complicated by the possibility of different evolutionary rates in different sequences. In sequences that do not encode TRIMCyp, approximately half of the region included in this analysis consists of coding sequence. In sequences containing the T allele, the entire region could be considered to be noncoding and thus potentially under relaxed selection. In these sequences, exon 8 is still used to code for TRIM5g; however, no biological function has been described for this isoform. Due to this uncertainty, no firm conclusions can be drawn from our phylogenetic analysis about the timing of the evolution of the T allele. Thus, the most parsimonious explanation for our data remains that TRIMCyp-related sequences evolved once in the ancestral Asian macaque lineage. Although it is unlikely that Old World TRIMCyp itself has been lost by lineage sorting, it should be noted that lineage sorting has almost certainly played a part in the evolution of this gene.