We were skipping SV deletion type variant if start >= end. The case start = end can be a valid case, for example, when SVLEN=1.
Fixed that check to see if start > end or if there is no SVLEN or END information.
Unit test fix:
AnnotationSource_File_VCF.t
The type is exact and the custom line is as follows -
21 25585735 del2 TG T 67 SEGDUP;RF
So it deletes the G at 25585736 position. Correct way to represent it in SVLEN is with POS=25585736, REF=T (always use the base before polymorphism happens) and SVELN=1 (the difference between ref and alt sequence).
Parser_VCF.t
The BND are case affected by the ensembl-io change because it does not have any SVLEN and END. The end here we will be now end=start. Before it was getting end=start+ref length-1. END position are not exactly clear for BND but the former ones seems more clear.
Depends on https://github.com/Ensembl/ensembl-io/pull/171
Skipping variant:
We were skipping SV deletion type variant if
start >= end. The casestart = endcan be a valid case, for example, when SVLEN=1. Fixed that check to see ifstart > endor if there is no SVLEN or END information.Unit test fix:
AnnotationSource_File_VCF.t
The type is
exactand the custom line is as follows -So it deletes the G at 25585736 position. Correct way to represent it in SVLEN is with POS=25585736, REF=T (always use the base before polymorphism happens) and SVELN=1 (the difference between ref and alt sequence).
Parser_VCF.t
The BND are case affected by the ensembl-io change because it does not have any SVLEN and END. The end here we will be now
end=start. Before it was gettingend=start+ref length-1. END position are not exactly clear for BND but the former ones seems more clear.