Thread by @Mike_Honey_

Earlier this month the XFL variant was defined, setting a new benchmark in SARS-CoV-2 evolution: a "quadruple-recombinant", meaning it’s evolution includes 3 recombinant ancestors.

It’s a stark example of how we are giving this virus every opportunity to evolve.
#COVID19 #XFL
🧵

The diagram summarizes it’s evolutionary journey, which includes 3 ancestor recombinants and several other sub-lineages before XEC.18 and XEU recombined to form XFL.
🧵

The start of the upper branch of this tree is the JN.1.11.1 variant, which evolved into the more successful KP.2.3 and KP.3.3 in early 2024.

All 3 variants were in circulation for around a year, with JN.1.11.1 peaking at 2% globally, with KP.2.3 at 9% and KP.3.3 at 8%.
🧵

Then the lower branch of this tree progressed with KS.1.1 (descended from JN.1.13.1) recombining with KP.3.3 to form the very successful XEC variant, around June 2024.
🧵

The KS.1.1 variant was quite insignificant, peaking at 0.2% globally. But it’s progeny XEC was very successful, rising to 23% itself, and spawning many successful sub-lineages.
🧵

Next the upper branch of this tree progressed with KP.2.3 recombining with XEC to form XEK, around August 2024.

The XEK variant has been quite insignificant, peaking at 1.2% globally. It did spawn a few child variants, but those were all even less successful.
🧵

Then the XEK variant recombed with XEC.8 to form XEU, around September 2024.

The XEU variant was very insignificant, peaking at 0.13% globally. It has not been detected since early March.
🧵

Finally, the XEU variant recombined with XEC.18 to form XFL, probably around November 2024.

The XFL variant has not been significant so far, rising to 0.11% globally.
🧵

Each recombinant presents a fresh arrangement of spike mutations, opening new evolutionary possibilities that work in combination with the step-wise mutation process. More on mutation and recombination in this paper:

🧵sciencedirect.com/science/articl…

To put the frequencies above in context, consider that with people being re-infected every 1-2 years on average, there are several billion infections each year. So a variant rising to just 1% for a month or so will likely have caused several million infections.
🧵

Still, it’s striking that there were points along the evolutionary path of XFL where just a small amount of transmission suppression would have reduced the odds of the two parent variants from meeting in a co-infected host.
🧵

But the current public health settings of unfettered spread and forced infection of the immune-compromised gives the virus every opportunity to try out these unlikely combinations.
🧵

A huge thanks to variant hunter Xu Zhou for the original diagram.

🧵

https://x.com/xz_keg/status/1910713021372260594

And further huge thanks to graphic designer Angie Cibis who volunteered to give the diagram a make-over for greater effect. Angie’s first draft (working from my rambling spiel) was quick and very very close -just one minor (and quick) rev to perfection!

🧵angiecibis.com

Interactive genomic sequencing dataviz, code, acknowledgements and more info here:

🧵 endsgithub.com/Mike-Honey/cov…