Hi Dr. Holmes,
We did the ML tree for the ZIKV
COMPLETE dataset and as you suggested there is a clade comprised by 8 sequences with anomalously long branches.
So, we inferred the nucleotide substitution rate of the ZIKV
NON-HUMAN E genes using two new datasets: 1) ZIKV
NON-HUMAN-A that included all ZIKV sequences retrieved from mosquitos and monkeys minus those 8 “anomalous” secuences; and 2) ZIKV
NON-HUMAN-B that included the same sequences than ZIKV
NON-HUMAN-A minus those previously characterized as recombinant in the original Faye et al. (2014) publication.
Despite the new rate estimations are between 1.0-1.5 ×10-3 , as Dr. Rambaut observed before, the interval spans more than an order of magnitude. Therefore, I assume that these sequences have poor temporary structure. On the other hand, I wonder if anyone knows which are the ancestral lineages of the recombinant strains characterized by Faye et al. (2014). I could not find this in that paper.
Below I place the Table with the clock estimates and the ML ZIKV
COMPLETE E gene tree.