View Single Post
Old 02-20-2019, 06:28 AM   #2
geejaytee
Junior Member
 
Location: UK

Join Date: Feb 2019
Posts: 2
Default

They do not need correcting - as the deamination (C>T and G>A) damage is random, the predicted allele frequencies for these (false) calls will be small (<0.25, say), and will be of low quality. A simple filter on quality and allele fraction can then be used to filter these out as they cluster separately to the true variants.
geejaytee is offline   Reply With Quote