AngieHinrichs
commented
11 months ago Thanks Shay -- the first 3 are consequences of branch-specific masking. The reversion A22995C is masked in BA.2.75, BA.5 and XBB because reversions on that base were too noisy. 241 (any change) is masked in everything from BA.2 onward -- that's interesting that it has mutated again C > T > A. In retrospect it probably would have been better to mask only the reversion T241C instead of all 241 changes, but... maybe too late now. Masking G19326A in all of XBB based on early observations was probably also a bad idea.
I'm finding that 21998 is super-messy all across XBB, so I might have to mask that too.
FL.4.4 All samples lacking 22995A and in the phylogeny it seems to be there https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_6dbf_51d490.json?c=gt-nuc_22995&label=id:node_6727233
FU.2.1 241 has T in the tree while 94% of the samples got A (which I think is weird as this is B.1 defining mutation) https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_17bcf_5265d0.json?c=gt-nuc_241&label=id:node_6634300
XBB.1.16.8 There seems to be reversion not recognized in the tree. the tree shows 19326 with G while only 5.5% of the samples has this mutation (1.7% has N In this position). https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_2135b_58cdf0.json?c=gt-nuc_19326&label=id:node_6623401
FL.10.1 The path to this variant include 21998A although one step after the defining branchpoint this mutations is reversed (and that’s the main branch of the variant) and indeed only 2.9% of the samples have that mutation in their VCF (less than 1% has N). https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_1b9c2_7995a0.json?label=id:node_6713858