Evolutionary origins of the SARS-CoV-2 sarbecovirus lineage responsible for the COVID-19 pandemic

Maciej F Boni; Philippe Lemey; Xiaowei Jiang; Tommy Tsan-Yuk Lam; Blair Perry; Todd Castoe; Andrew Rambaut; David L Robertson

The key difficulty in inferring reliable evolutionary histories for coronaviruses is that their high recombination rate (Graham and Baric, 2010; Su et al., 2016) violates the assumption of standard phylogenetic approaches because different parts of the genome will have different origins. In order to begin characterizing any ancestral relationships for SARS-CoV-2, non-recombinant regions of the genome must be identified, so that reliable phylogenetic reconstruction and dating can be performed. Evolutionary rate estimation can be profoundly affected by the presence of recombination (Schierup and Hein, 1999) . As there is no single accepted method of inferring breakpoints and identifying clean sub-regions with high certainty, we implemented several approaches to identifying three classic statistical signals of recombination: mosaicism, . CC-BY-NC 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.30.015008 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.30.015008 doi: bioRxiv preprint 17 / 25 phylogenetic incongruence and excessive homoplasy (Posada et al., 2002) . Our most conservative approach attempted to ensure that putative non-recombining regions had no mosaic or phylogenetic incongruence signals. A second approach was conservative with respect to breakpoint identification, while a third approach attempted to minimize the number of regions removed while still minimizing signals of mosaicism and homoplasy. The origins we present in Figure 5 (NRR1) are conservative in the sense that NRR1 is more likely to be non-recombinant than NRR2 or NRA3. As the estimated rates and divergence dates were highly similar in the three data sets analyzed, we conclude that our estimates are robust to the method of identifying the genome's non-recombinant regions.