1/ My take on the raccoon dog WNV report. Not that anyone needs another, but I read it, so might as well.

Open reanalysis of genomic data with different groups/perspectives and skillsets stimulates discussion and allows richer learning from limited data. This is good 2/
Only new data will help us move the needle on this discussion, and most of these data (if they exist) are in China. Given where the world seems to be headed (new Cold war?), I am not hopeful these data will be forthcoming.

1/ It's been a while since I updated the selection analysis of SARS-CoV-2 spike. The granularity of lineages is such that within-lineage analysis makes no sense, since any new persistent mutation (almost) now leads to a new lineage designation observablehq.com/@spond/ba1-sel… 2/ Within older lineages (BA.4, BA.5, etc, even BA.2.75) the pattern of "selection" one sees is primarily due to reversions, which could be real or could be sequencing errors. So it wasn't particularly interesting.

1/9 What happens during superinfection with different SARS-CoV-2 strains (hint: recombination)? A new preprint with @DarrenM98230782, Joel Wertheim, @nekrut and New York City (NYC) Department of Health and Mental Hygiene medrxiv.org/content/10.110… 2/9 CoV including SARS-CoV-2 are prolific at recombination. But if two genetically similar strains recombined, they leave little detectable trace. In this case, an individual was super infected by Alpha and Epsilon, which have enough differences.

1/18 "Selection analysis identifies unusual clustered mutational changes in Omicron lineage BA.1 that likely impact Spike function" biorxiv.org/content/10.110… with @DarrenM98230782 @jbloom_lab @Tuliodna @SpyrosLytras @robertson_lab and many others. 2/18 13 of the S-gene mutations in BA.1 (Omicron) fall into three "unusual" clusters: sites that have been subject to **negative** selection or conserved in other SARS-CoV-2 lineages and closely related bats and other human beta-CoV (blue below)

1/18 Since a certain high ranking government official thrust the topic of "prediction" into the limelight.

Good news: we **can** make useful predictions (see below)

Bad news: General prediction of viral evolution (PVE) is fundamentally an intractable problem. 2/18 Saying that new variants will emerge is both profoundly true and profoundly useless. It's like saying that a market crash/bear market is coming sometime in the future with no specifics. A useful strategy would be to develop hedges based on the best available information.

1/6 We developed a new #omicron S-gene selection dashboard (will keep updating daily) using @GISAID sequences with UI in @observablehq and analysis by @hyphy_software
observablehq.com/@spond/ba1-sel…
@aglucaci @DarrenM98230782 @stvnwvr 2/6 Selection pressure on #Omicron seems to have *stabilized* both in terms which sites are under selection and the overall dN/dS estimates, where dN/dS on internal branches is dropping down where the leaves are (dN/dS internals > dN/dS leaves was very unusual)

1/4 2012/12/16 #Omicron (BA.1) selection (S-gene) detection update using >6800 sequences @GISAID
1. New + selection signal at S/1081 (I1081V)
2. New - selection signal at S/60 (TCC=>TCT)
3. Some of the clade defining sites are off the selected list (373,864,969,981) 2/4 S/1081 is perfectly conserved (I) in sarbecoviruses (@SpyrosLytras), and "V" was found in ~100 isolates prior to #Omicron. It is at ~5% in BA.1 sequences.

1/4 2021/12/15 #Omicron selection (S-gene) detection update using >5200 sequences @GISAID
1. New selected site S/V193L
2. Continued selection on most clade defining site
3. Clear pattern of epistasis between S/417 and S/452
4. First site showing clear *negative* selection S/1146 2/4 S/V193L appears in sequences from Denmark and Israel. It has been previously positively selected in global data in Aug/Sep 2021. Seen in a 100s of sequences before (B.1.1.7 mostly). The site in Sarbecoviruses has a few distantly related Ls. The IVL set is quite exchangeable

1/13 Dec 13th #omicron selection update using >3200 BA.1 sequences from @GISAID. Subject to limited data/QA issues.
• ~12% of S-gene along internal tree branches is evolving with dN/dS > 1.
• Continued selection on phenotypically relevant sites (S/346, S/452). 501, 701 as well. 2/13 There's currently a weird inversion of selective forces: dN/dS is higher on internal branches than on leaf branches. We expect the opposite (e.g., academic.oup.com/mbe/article/24…) for most RNA viruses, as leaf branches carry some deleterious mutations.

1/8 @DarrenM98230782 and I summarize our thoughts on the near-term evolutionary fate of the #Omicron lineage and its individual mutations
(virological.org/t/a-fool-s-err…). As the title implies, prediction of viral evolution has not historically been successful. 2/8 Anyone who claims to *know* where this is going is a fool. The emergence of #Omicron testifies to it. We consider four scenarios for the lineage in the near/midterm. This is also purely from an evolutionary perspective; the ultimate clinical outcomes are a separate story.

1/8 Selection analysis update on the S-gene of #Omicron. Caution: preliminary data and analysis; based on ~800 sequences (12/7) from @GISAID classed as B.1.1.529 by #Pangolin. There appears to be signal of positive selection in circulating #Omicron sequences. @DarrenM98230782 2/8 Thanks to all the data contributors, including @Tuliodna and SA colleagues! We find evidence of positive selection at S/346 (K), S/701 (V), S/211 (several states, possibly due to indel alignment issues). Genomes have a lot of noise around S/214, so this could be an artifact.

1/18 "Selection analysis identifies significant mutational changes in Omicron that are likely to influence both antibody neutralization and Spike function" virological.org/t/selection-an… @DarrenM98230782 @Tuliodna @jbloom_lab @robertson_lab @nekrut @LemeyLab and many others. 2/18 There are 14 mutation in #omicron S-gene that have been under negative selection (or neutral evolution) prior to Nov 2021. This pattern was NOT seen in previous VOC where many of the signatures sites had been detectably selected prior to emergence cell.com/cell/pdf/S0092…

1/4 In a May 2021 preprint on the common evolutionary trajectories of human beta-coronaviruses, we (@EvolveDotZoo, Marina Escalera-Zamudio and others) identified four sites, including S/796 found in #omicron that we hypothesized might be involved in human adaptation 2/4 In particular, S/796 has experienced what we termed "stepwise evolution" in SARS-CoV-1 and is near the trimerization surface, which undergoes conformational rearrangements during viral fusion

1/11 Can the evolutionary history of sarbecoviruses help predict the effect of mutations in #omicron? Experimental measurment of phenotypic effects is the gold standard (e.g. see the magnificent DMS-based predictions by @jbloom_lab). What about evolutionary predictions? 2/11 Obviously, if a mutation has been observed at appreciable frequencies in SARS-CoV-2 circulation, this provides evidence that it is not particularly deleterious or may be adaptive (at the time it was circulating, anyway).

There's definitely a strong signal of selection on Spike in #Omicron compared to reference clades in our preliminary RASCL analysis of ~60 sequences (thanks @aglucaci, more coming)
1. Spike is under positive selection
2. Spike is under stronger selection than background There are 9 spike sites where there's stronger selection in #Omicron compared to other clades according to Contrast-FEL (academic.oup.com/mbe/article/38…). Sorted by q-value here (stronger evidence at the top)

#SARSCoV2 selection analyses updates. We switched to running sliding windows analyses (blocks of 3 months) to deal with data volumes and get temporal trends. The current state of analyses is at observablehq.com/@spond/selecti… This includes an at-a-glance view of selection profiles on the most recent time window

1/ A recent preprint (papers.ssrn.com/sol3/papers.cf…) reporting detection of sequence and antibody evidence for SARS-CoV-2 in Italy in the fall of 2019 presents results that are at odds with the current early SARS-CoV-2 timeline. 2/ It may be tempting to dismiss these results as false positives or some other data artifact (e.g.

https://twitter.com/MichaelWorobey/status/1424483875384958981

), but should it be done for these “inconvenient" data?

An important bit of forensic bioinformatics by @jbloom_lab
biorxiv.org/content/10.110… The analysis of recovered sequences does not fundamentally change our current understanding of early SARS-CoV-2 evolution, but it does make the hypothesis of a single-source wet market outbreak implausible.

An update on #SARSCoV2 selection analysis using @GISAID data (observablehq.com/@spond/natural…). I added a simple 5-category classification for each potential interesting site. One category = one point. The more points, the more interesting a site is. Category 1. Is the site under selection using statistical comparative methods?

Category 2. Is there a large (>20%, which is incidentally what you can detect with mixed bases) fraction of minority alleles (synonymous or non-synonymous) among viral haplotypes at the site.