Closed priesgo closed 2 years ago
Original request was referring to https://github.com/BU-ISCIII/viralrecon/issues/14 and https://github.com/andersen-lab/ivar/issues/104. The first one is related to variants in overlapping read pairs, while the second one is related to overlapping variants.
Starting with the second, the issue of overlapping variants is dealt with differently in CoVigator. We annotate the variants with VAFator and then only use for the consensus sequence those with VAF >= 0.8, we do not use ivar consensus
but bcftools consensus
. In any case, the overlapping variants reported in the ivar issue look funny, isn't that an insertion and a SNV? If that is the case they do not overlap.
About the first of overlapping reads. This issue of different Phred qualities is reported on samtools 1.8, we use 1.12. I went through the samtools release notes here https://github.com/samtools/samtools/releases and could not find any reference to this issue being fixed. In any case in #37 I added several options to add optional arguments into the variants calling process such as --args_ovar_samtools='--ignore-overlaps'
Spin off feature request from #28