search

neherlab / ncov-simple

2 stars 1 forks source link

Consider not hard masking S:95 #30

Closed corneliusroemer closed 2 years ago

corneliusroemer commented 2 years ago

S:95 does show a lot of reversions but the problem is not as severe as with S:142.

Unmasking S:95 could help reduce size of the main Delta polytomy

@rneher what shall I try? Soft masking S:95 or not masking at all? If we soft mask, it'll cause ladderization since S:95 is not used for tree building. So really we should probably try not masking at all, not even soft.

rneher commented 2 years ago

yes. soft masking makes sense for recurrent mutations that often are part of big clades together with other mutations. Masking them for topology reduces the risk of tree builder pulling clades together that shouldn't be together. Hard masking is for sites that would generate artificial clades by pulling together sequences that have sequencing errors.