Smoove_B wrote: Tue May 05, 2020 12:25 pm
More perspective from Harvard Public Health Professor:
https://twitter.com/BillHanage/status/1 ... 6436613126
This preprint has been getting attention. It claims that the SARS-CoV-2 virus is mutating into a more transmissible form as the pandemic wears on. I think those claims are suspect, to say the least
I read that thread this morning. I read the paper and can't map this statement to any particular claim in the paper. The claim they seem to be making is the one in his 2nd tweet. The major argument was the 2nd strain became predominant. And the data does seem to bear that out which he actually talks through in the thread. 1 is implied by 2 but they don't make the direct claim. They instead say it should be explored as a possibility. At least, that is how I read the paper.
https://twitter.com/BillHanage/status/1 ... 6244247553
From the discussion in the paper - I don't know how you read the below and then make the statement he makes in the 1st tweet:
When we embarked on our SARS-CoV-2 analysis pipeline, our motivation was to identify mutations that might be of potential concern in the SARS-CoV-2 Spike protein as an early warning system for consideration as vaccine studies progress; we did not anticipate such dramatic results so early in the pandemic. In a setting of very low genetic diversity, traditional means of identification of positive selection have limited statistical power, but the incredibly rich GISAID data set provides an opportunity to look more deeply into the evolutionary relationships among the SARS-CoV-2 sequences in the context of time and geography. This approach revealed that viruses bearing the mutation Spike D614G are replacing the original Wuhan form of the virus rapidly and repeatedly across the globe (Fig. 2-3). We do not know what is driving this selective sweep, nor for that matter if it is indeed due the modified Spike and not one of the other two accompanying mutations that share the GISAID “G-clade” haplotype. The Spike D614G change, however, is consistent with several hypotheses regarding a fitness advantage that can be explored experimentally. D614 is embedded in an immunodominant antibody epitope, recognized by antibodies isolated from recovered individuals who were infected with the original SARS-CoV; this epitope is also targeted by vaccination in primate models (Wang et al., 2016). Thus, this mutation might be conferring resistance to protective D614-directed antibody responses in infected people, making them more susceptible to reinfection with the newer G614 form of the virus. Alternatively, the advantage might be related to the fact that D614 is embedded in an immunodominant ADE epitope of SARS-CoV (Wang et al., 2016), and perhaps the G614 form can facilitate ADE. Finally, the D614G mutation is predicted to destabilize inter-protomer S1-S2 subunit interactions in the trimer, and this may have direct consequences for the infectivity of the virus (Fig. 4). Increased infectivity would be consistent with rapid spread, and also the association of higher viral load with G614 that we observed in the clinical data from Sheffield, England (Fig. 5).
Each of the ways we anticipated we might find evidence of positive selection in Spike are being manifested among subset of the sites that are accruing mutations. While the D616G mutation is the only one that is dramatically increasing in frequency globally (Fig. 2-3), the L8V mutation may be on the rise in the local epidemic in Hong Kong (Fig. S7). To date, mutations are extremely rare in the Spike RBD, but the mutation G476S is directly in an ACE2 contact residue. The mutation L5F occurs in many geographic regions in many distinct clades, suggesting it repeatedly arose independently, and was selected to the extent that was frequent enough to be resampled. Finally, the mutation S943P seems to have been transferred by recombination into diverse viral backbones that are co-circulating in Belgium (Fig. 6); we also found strong evidence of recombination in other regional sample sets (Fig. S8). Recombination among pandemic SARS-CoV-2 strains is not surprising, given that it is also found among more distant coronaviruses with higher diversity levels (Graham and Baric, 2010; Li et al., 2020; Rehman et al., 2020). Still, it has important implications. First, recombination cannot be detected without simultaneous coinfection of distinct viruses in one host. It is not clear if such co-infections might be happening prior to the adaptive immune response, or in series with reinfection occurring after the initial infection stimulated a response. Recombination may be more common in communities with less rigorous shelter-in-place and social distancing practices, in hospital wards with less stringent patient isolation because all patients are assumed to already be infected or in geographic, or in regions where antigenic drift has already begun to enable serial infection with more resistant forms of the viruses. Also, recombination provides an opportunity for the virus to bring together, into a single recombinant virus, multiple mutations that independently confer distinct fitness advantages but that were carried separately in the two parental strains.
Tracking mutations in Spike has been our primary focus to date because of the urgency with which vaccine and antibody therapy strategies are being developed; the interventions under development now cannot afford to miss their contemporary targets when they are eventually deployed. To this end, we built a data-analysis pipeline to explore the potential impact of mutations on SARS-CoV-2 sequences. The analysis is performed as the data becomes available through GISAID. Experimentalists can make use of the most current data available to best inform vaccine constructs, reagent tests, and experimental design. While the GISAID data used for the figures in this paper was frozen at April 13, 2020, many of the key figures included here are rebuilt each day based on the newly available GISAID data. While our initial focus is on Spike, the tools we have developed can be extended to other proteins and mutations in subsequent versions of the pipeline. Meanwhile understanding both how the D614G mutation is overtaking the pandemic and how recombination is impacting the evolution of the virus will be important for informing choices about how best to respond in order to control epidemic spread and resurgence.
The more interesting discussion is that others have ascribed this to a systematic sequencing error versus true mutation. That sounds fascinating from a systems perspective. In effect, how is everyone doing the same wrong thing?
https://twitter.com/EBIgoldman/status/1 ... 5800629249
