Thread by @ewanbirney

Some COVID thoughts on this bright, beautiful Christmas Eve morning in London.

Context: I am an expert in genetics and computational biology; I know and chit-chat with experts in viral phylogeny, infectious epidemiology, immunology + testing. I have a COI that I am a consultant to Oxford Nanopore that make a COVID test. I am also on the AZ vaccine trial

Reminder: SARS_CoV_2 is an infectious virus which causes a nasty disease in a subset of people, often leading to death. If we let the virus go through the population not only would many people get this disease, but also healthcare systems would be overwhelmed.

Just recently we have learnt of at least two new strains of the virus which increase transmission, one first discovered in South East England, one first discovered in South Africa.

When viruses jump species they often are not optimised for their new species, and so there is a strong selection pressure to increase transmission; indeed, it is possible that the strain that ended up dominating European transmission in May was slightly better transmitter

(Viruses which causes widespread deadly diseases often have a selection pressure to become less deadly, but the SARS_CoV_2 situation is different because it is only a subset - more old, male, overweight - of humans that have a fatal disease).

Working out whether it is different (and higher) in transmission is easier than quantifying that difference - multiple disparate lines of evidence point to the B1.1.7 strain (South England discovery) having transmission advantage from genomics through to epidemiology.

Quantification relies on a model of transmission and it is complex, but my favoured way to think is that it is adding 0.5 to 0.9 to R (the reproduction rate); note, even the word "adding" (implying a linear model) is up for debate in how to model this.

The B1.1.7 strain has been seen and confirmed via genomics in Denmark, Netherlands, Italy and Australia in small numbers, but the lack of broad surveillance across many countries means this is likely an underestimate in countries and worldwide.

It is also pretty inevitable that we will have other transmission selection variants in the future and we will need our A-game on surveillance for immune escape as vaccination kicks in. Time for everyone to tool up.

A quirk of the strain is that it causes the RT-PCR test primers from one set to "drop" one primer set (from 3); this does not harm the test robustness but is a proxy to track this. This "S gene drop out" (or "ORF1a and N gene only positive") has more numbers and more places.

There is some indicative evidence of this "S gene drop out" in the US at least, and other countries presumably are analysing the data. S gene drop out happens also on other lineages, so one needs to do genomic sequencing to confirm

It's imperative in my view that many countries increase their genomic surveillance of the virus (I note my COI here to Nanopore which makes one of the sequencing machines that can do this - this a good thing, but better to be transparent!)

Looking at the spread of the S-gene drop out around the UK (recent ONS survey), the US S-gene drop out increase and the depth of the placement of the Netherlands+Italian samples into the B1.1.7 lineage, I strong suspect the B1.1.7 lineage is in many other locations. (Bugger).

Given it was hard to contain the SARS_CoV_2 strain outbreak in March - though we did not know all the details then and the concern - I think it is going to be hard to contain this spread sadly, but worth trying and every bit of containment will buy time...

And that time will be useful. Thankfully - and my god it would be depressing if it was not the case - we have 2 vaccines with trials and regulatory approval (BioNTech/Pfzier + Moderna) and one with interim results and submitted to regulators (at least in the UK, Oxford/AZ).

There is every reason to think these vaccines will work against the new strains - though one should check (I am sure people are!) - new strains and tweaks emerge all the time and the trials of the vaccines were against a mixture of "naturally occuring" strains.

Furthermore all 3 vaccines elicited T-cell as well as B-cell response, and T-cell response is more robust against variation (T-cell 'sense' fragments, and usually many different fragments). Some T-cells are command-and-control of B-cells, and allow robust B-cell response

So - where does this leave us? For all countries one needs to suppress transmission of the virus until the population is vaccinated enough to limit acute disease and ideally help slow transmission itself (thus allowing less restrictions).

For the countries where the new strains are present at a large rate - at least UK, South Africa, but possibly more this means more restrictions. The positive is that vaccination is happening, and all wind to their sails.

For countries without evidence of the new strains, the strategy remains the same but with the added need to spot any potential new strains - either from UK or SA or potentially inherent. Response on the new strain presence is going to be complex.

There is a huge tendency to draw national - or regional - boundaries around this - for good reason in that many responses have to come from governments and those boundaries are where governments control change. However, this national view is not how the virus behaves (obviously)

Furthermore although doing comparisons between countries is super-useful (I still ... commend the Japanese backtracing model that leverages small scale super-spreading events for suppression) comparisons can easily spill over to unhelpful blame games

Did the UK "cause" the new strain to emerge due to high levels of the virus? Well... many countries have equally high levels, some by design or not. Would it have been better for the UK not to do surveillance and be open about the results? Clearly not.

This struggle is between humans and the virus. All other groupings of humans are constructions of humans which we have to understand (they are the reality we operate in) but are not boundaries the virus respect. We have to solve this as humans, together.