IVDC-HB-04/2020is also suspect - it has 5 non-synonymous mutations and 3 synonymous
The nextstrain tool-tip is misleading here. The reference used has over-lapping annotations ORF1a and ORF1ab. There is a total of 3 mutations inferred for this branch. C1023T, C1025T, A18460G
The first two change the aa sequence of ORF1a in coding 253 and 245 (and these are the same as the mutations listed in ORF1ab).
I was mistaken. This is wrong:
The last mutation is synonymous also in ORF1ab after the slippage site.
So: 2 adjacent non-synonymous, 1 synonymous.
Correction:
The last mutation at A18460G is also non-synonymous. All three mutations are non-synonymous