neherlab / ncov-simple

2 stars 1 forks source link

Private mutation counts are very high ~30 for regional lineages in countries that don't submit to Genbank #20

Open corneliusroemer opened 2 years ago

corneliusroemer commented 2 years ago

The private mutation filter can't really be used very tightly at the moment because certain diversity that never made it to Genbank submitters (UK, Germany, US, Switzerland) is not part of current Nextclade placement trees.

Perfectly fine sequences in countries like Indonesia can easily get private mutation counts of 30.

So private mutations can only be used very insensitively within diagnostic.py unless Nextclade uses a GISAID reference tree.