Thread by @SolidEvidence

Time to update my lineage cheat sheet.

This virus is doing some strange things.

1/

Other than the BA.2.86/JN lineages, the hottest trend right now is for lineages to pick up the FliP mutations L455F/F456L in addition to A475V. It's pretty clear that this combination is more antibody resistant.
2/

The main lineages with this combination right now are JD.1.1, GK.3.1, FL.15.1.1 and GW.5.1.1.
3/

You would think the most prominent lineages would be derived from one of the very prominent lineages that was only one mutation away such as JG.1, HK.3, or JF.1. But that is not the case.
4/

Almost all of these lineages with the trio are derived from lineages that branched off earlier in the tree and recreated the whole trio of mutations from scratch.
5/

The most prevalent of these trio lineages is JD.1.1, which is derived from a little known lineage (XBB.1.5.102) that circulated mostly in Brazil. In the early summer it independently picked up 455F, 456L, and finally 475V.
6/

JD.1.1 mostly stayed put in Brazil until about July, when it moved into North America and Europe, and it has been expanding at a steady rate since then.
7/

Although JD.1.1 is the most prevalent of the trio lineages, the one I’m keeping my eye on is GW.5.1.1.
This lineage independently acquired 486P, 455F, 456L, (478I), and 475V.
8/

In addition, @LongDesertTrain recently noted that GW.5.1.1 has completely deleted Orf 7 and 8. (Something even the sequencers had not noted). It isn't the first lineage to do this, so I'm curious to see what it will do for the virus.

9/

https://twitter.com/LongDesertTrain/status/1718769959688360206

Not sure if anyone has noted this already, but the A475V + FliP mutations are showing up almost exclusively (>99%) in lineages lacking S:478R. I’m guessing 475V and 479R are incompatible. We’ll see if that is true.

10/

As always, let me know if I got any of my facts wrong.

11/11