Thread by @ewanbirney

COVID thoughts on a cloudy late January day in London - longer evenings, and our house lockdown rhythm has settled somewhat.

Context: I am an expert in human genetics and computational biology; I know experts in infectious epidemiology, viral genomics, clinical trials and immunology. I have some COIs: I am a long established consultant to Oxford Nanopore and I am on the Ox/AZ vaccine trial.

Reminder: SARS-CoV-2 is infectious virus which causes a severe disease in a subset of people (older; more obese; male risk factors) often leading to death. If we let the virus go through the population both a large number of people would die and healthcare systems would melt

(the resulting likely dysfunctional healthcare systems would then mean yet more people with COVID - and other diseases - would die or have debilitating diseases).

England got close to this meltdown level in January due to exponential rise in transmission in December due to B.1.1.7 variant (more below); as multiple places have now shown, one cannot manage an "acceptable rate" of entry into healthcare - epidemics can't be "tuned" >>

Epidemics come in two forms; growing or shrinking. If you are growing it will overwhelm your healthcare system at the moment. If you can make shrink, simply make it shrink to low levels and there is v. little benefit in stopping the shrinking early.

Remarkably we now know many vaccines are safe and prevent disease and infection; the protective effect on disease far more than infection, but it is easier to measure in a trial the protective effect on infection.

As expect, the virus has also mutated (which it does all the time) and some of those mutations have enhanced transmission, disease severity in some cases and perhaps immune suppression in other cases.

These specific lineages are given defined identifiers to track - B.1.1.7, B.1.351 and P.1 but often known from the location they were first discovered - "English", "South Africa" and "Brazilian" variants

Countries with good genomic surveillance can spot variants with new biology earlier (as the UK did with B.1.1.7) or track their progress more accurately (as Denmark is doing with B.1.1.7). All countries should ideally improve their genomic surveillance.

(As a note: we need to share the data openly and fast, including raw data, of the genomic information; please sign our petition for fully open data sharing here: covid19dataportal.org/support-data-s…)

The faster transmission of B.1.1.7 and other variants in Europe has changed the ability to suppress SARS-CoV-2 transmission; currently only stronger NPIs (lockdown rules) have suppressed B.1.1.7 - which have worked in UK, Ireland, Israel and it looks like turning now in Portugal.

The exit seems to be vaccines, and thankfully the vaccines looks like they work at least against B.1.1.7 (extensive lab data. Best "real world" data from Israel at the moment; I would hope to see some from the UK).

It might be that there are some clever tricks to play on TTI; some finessing on NPIs; Summer will help for sure; but the main thing is vaccination rates - rates of production; rates of supply; rates of application.

This has blown up in the EU with plenty of politics, and a spice of how UK has structured vaccine supply but politics and blame games wont change the fundamentals about rates - one needs to improve all these rates. That is the core issue. Basically little else matters here.

Looking beyond UK and EU perspective, the real unit of population to worry about is the world. It is great to see the large Serum Institute in India pumping out CovidShield (Ox/Az) vaccine, and I think all the Chinese vaccination pushes needs a second look at.

Furthermore we cannot assume that the vaccination will work well for all strains. It *might be* that any SARS-CoV-2 vaccination calms down the autoimmune response (this is "optimistic scenario") but we have to plan for the worse case - vaccine escape in 2021

Although depressing and a bit exhausting to think about, we now know that mRNA vaccines work, and these are the easiest to tweak in synthesis and I think for multiple vaccinations (Ox/AZ could be repositioned but the multiple adenovirus infection might get in the way)

(I'm beyond my expertise window here, so love to hear some vaccine experts opinions on this. I appreciate it is fearsomely complex).

In the meantime, despite lockdowns we need to get all our tools sharpened up for 2021 - better TTI, better surveillance (genomics), data integration, international genomic sharing. I am convinced 2021 will be better than 2020, but it is a long road.

Finally, as ever, this is a humans vs virus conflict. All humans are one unit here.