Thread by @Mike_Honey_

With the LP.8.1.* variant on the way to dominance in most places, it is time to ponder which variant might drive the next wave.

The leading contenders at this point are LF.7.7.2, NB.1.8.1, XFG and XFJ.

#COVID19 #LF_7_7_2 #NB_1_8_1 #XFG #XFJ
🧵

I show them above using a log scale, so you can compare their growth rates vs the most common LP.8.1.* sub-lineage: LP.8.1.1. Note the recent sample volumes are quite low, so the right side of this chart might not be a representative picture.
🧵

LF.7.7.2 is descended from FLiRT JN.1.16.1. LF.7 added several Spike mutations: T22N, S31P, K182R, R190S and K444. Then LF.7.7.2 added the Spike H445P mutation.
🧵

LF.7.7.2 has been most successful in Canada and the US, rising to 8% frequency in both countries. The overall growth rate has slowed, but it still looks quite steady in the US.
🧵

NB.1.8.1 is descended from XDV.1.5.1. XDV was a recombinant of XDE and JN.1. XDE was a recombinant of GW.5.1 and FL.13.4, so this represents the last current variant with any non-JN.1 ancestry.
🧵

XDV.1 added the F456L mutation, then XDV.1.5 added G184S and K478I. NB.1 then added Spike mutations: T22N and F59S. Then NB.1.8 added the Spike Q493E mutation that characterised KP.3.1 FLuQE – an example of convergent evolution. Finally NB.1.8.1 added the A435S mutation.
🧵

NB.1.8.1 had mainly been reported from Hong Kong, rising to 199% frequency. It has also started to show signs of growth in several other countries in the region, plus the US - all adding to its credibility.
🧵

XFG is a recombinant of LF.7 and LP.8.1.2, with a presumed origin in Quebec.

Growth of XFG accelerated in the recent French samples to 7%. Growth in the US reached 5%, with most reported from New York and New Jersey.
🧵

XFJ is a recombinant of LF.7 and LS.2. LS.2 was descended from JN.1.18.5.

Starting from February, XFJ has been most successful in Italy, rising to 9% frequency.
🧵

XFJ's growth to 2% in the US has been driven by recent samples from New York - those collected in April were not collected from travellers.
🧵

Here’s a thread by variant hunter Federico Gueli, who first spotted what is now designated as XFJ. He highlights some of the interesting convergent evolutionary steps that helped it along it’s journey.

🧵skyview.social/?url=https%3A%…

Note the consensus seems to be forming among the variant hunters that what has been designated as XFJ might in fact be 2 closely-related but distinct recombinants.

🧵github.com/cov-lineages/p…

Here’s a thread by variant tracker Andrew Urqhart (maintainer of the fabled “Collection 42” on cov-spectrum), tracking the spread of XFJ sample-by-sample:

🧵skyview.social/?url=https%3A%…

I will stick with NB.1.8.1 as the leading contender. It is showing growth in multiple countries. XFJ is chasing hard, with quite strong growth in early April.

I will continue to monitor this topic.
🧵

The usual caveats apply - recent sample sizes are smaller which might skew these results, and “global” sequencing data is dominated by wealthy countries, with many under-sampled regions.
🧵

LF.7.9, NY.9 and XEC.25.1 have been eliminated from this analysis, as they each showed unimpressive recent growth and limited geographic spread.
🧵

Here’s the overall variant picture for New York, from March to early April. The sequencing volumes seem quite steady at around 50 per day, which probably gives a fairly representative view.
🧵

LP.8.1.* continued to extend it’s domination, up to 59%. JN.1.* +FLiRT (LF.7.7.2, LF.7.9 et al) seems stalled at around 20%. XFG and XFJ look the most dynamic.
🧵

Huge thanks to Federico Gueli for his tips on new lineages to watch out for, eg

🧵skyview.social/?url=https%3A%…

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

🧵 endsgithub.com/Mike-Honey/cov…