Dear #COVID19 twitter, a short respite from the bad news on the epi side, to share neutral results from the genomics side. Coronaviruses are highly recombinant; this means identifying evol origins is challenging, as different parts of the genome can have different origins. 1/
We used 3 approaches to identify non-recombining regions of the genome. Here is an example w the peaks in this graph showing the degree of support for a recombo breakpoint at that nucleotide position (x-axis). The A'BC region is one of our candidate non-recombining regions. 2/ 
Some individual recombination events could be identified, esp including two bat (R. sinucus) sequences from Zhejiang province (2015,2017), as well as a short region in the receptor binding domain of SARS-CoV-2's Spike protein which a bat sequence (RaTG13, sampled in Yunnan.. 3/
... province in 2013) seems to have acquired through recombination; the parental sequence here is not available. #SARSCoV2 itself does not appear to be a recombinant of any available sequences. 4/
The rationale for all this recombination analysis was to identify non-recombining regions which could be used for phylogenetic dating. Dating results below are for our most conservatively identified non-recombining region. 5/
We calibrated the evolutionary rate of #SARSCoV2 using two separate priors based on previously inferred rates for MERS and OC43. Results were robust to the selection of this prior. Divergence time of SARS-CoV-2 from its closest bat sequence (RaTG13) is about 40-70 years ago. 6/
Note that the pangolin lineages lie outside the RaTG13/SARS-CoV-2 lineage. We did not find that pangolins were necessary as an intermediate host. However, both pangolin and bat sampling is sparse around the SARS-CoV-2/RaTG13 lineage. 7/
This also means that there is substantial unsampled diversity in the sarbecoviruses (subgenus containing SARS & SARS-CoV-2) that has been circulating in bats for many decades. Many of these viruses have the adaptations/exaptations to bind the ACE2 of humans and pangolins. 8/
Thanks to a great team with expertise in recombination analysis, evolutionary rate calibration, and phylogenetic dating: @LemeyLab, @robertson_lab, Tommy Tsan-Yuk Lam, @john_jxw, @arambaut, @blairperry614, Todd Castoe. Paper is here: biorxiv.org/content/10.110…
