The actual evolutionary rate is not actually that important or interesting. The actual value tells you nothing about adaptation - there are too many competing stocastic factors that effect it. Sure, it might be different in a human-mosquito-human cycle than a monkey-mosquito-monkey cycle (or whatever the sylvatic host is) but it doesn’t mean it has changed in nature or is adapting.
The rate is just useful to calibrate the tree to infer the timescale - and getting a good reliable timescale on this outbreak seems like an important goal right now. It is best to calibrate the tree as close to the nodes that you are interested in. To understand when and how it got into South America, what we need is a couple of hundred genomes from affected countries spread over the last five years. They need to be directly sequenced from samples.
Completely agree that resequencing of older strains would be much better than trying to make back calculations from T/C rates. That’d require somebody having the original samples in a freezer though, as opposed to vials with already passaged strains. Hopefully it’ll be possible to get these - would certainly be very helpful!
It would certainly be interesting from an ecological point of view, it doesn’t seem like high priority. Sequencing the shit out* of samples in Asia, Polynesia and the Americas both contemporary and going back 5-10 years for frozen samples is what is needed right now. That would produce a good estimate of the recent timescale and provide a strong framework for tracking forward.
A formal scientific expression derived from ‘The Martian’
From these analyses and comments posted maybe we can conclude that African ZIKV sequences currently available, that were mainly retrieved from sylvatic transmissions, should not be used to calibrate/infer the timescale of recent human outbreaks because: 1) viral substitution rates in sylvatic and urban transmission cycles could be different, 2) most of the old African sequences were sequenced after multiple passages, 3) many recent African sequences could have problems of sequence integrity and/or recombination, or 4) a combination of all those factors previously described.
Because ZIKV outbreaks in South Pacific and the Americas seems to be of recent origin (2013-2015), older ZIKV sequences from urban epidemics (5-10 years ago) are probably only available in frozen samples from Asia and/or Western Pacific. Anyway, it could be also a good idea to performed retrospective studies of ZIKV infections in frozen acute febrile illness samples in the Americas taken before 2015 and that were negative for DENV and CHIKV. ZIKV may have been circulating for some time in the Americas, but remained undetected as ZIKV infections may have been attributed to other endemic arboviruses with similar clinical presentations.
Good Morning - In the spirit of Eddie’s concerns about isolate integrity and to supplement what Kristian posted - this is my catalog of sequencing as of 29Feb16 (still updating) via my digging through Genbank and going back to the original publications. We’re prepping a manuscript and this will in supplemental in CSV format but I thought you might find it useful? You will some sequences repeated (genome, envelope). Some sequences were only E gene, some were both so you’ll see duplication of accessions, they are in order of accession. For passaging it’s host followed by a number that number is how many times. Ie. Ae_psuedoscutellaris_7 = Ae. psuedo 7 times - Cheers, Mel