End position occurs before start position

[zhemingfan]

Hi @friend1ws ,

I've noticed that in the final output we received a call similar to: chr6    31427745        d_1188  A       <INS>   .       Too_small_size  END=31427708;SVTYPE=INS;SVINSLEN=0;SVINSSEQ=    TR:VR   107:8   61:0 Is there a possibility to add a check to coerce the end positions to be equal or greater than the start position for insertions and deletions?

Thanks

[kojix2]

Thank you for sharing your experience.

I encountered the same issue where the Circos Plot could not be displayed in the SV inspector of jbrowser2 due to the following record:

chr7    61067874    d_338   T   <INS>   .   Too_small_size  END=61067865;SVTYPE=INS;SVINSLEN=14;SVINSSEQ=TGAATCAAATGGTG TR:VR   60:16   23:0

As shown in the screenshot below:

When I ran bcftools view tumor.nanomonsv.result.vcf, I received the following warning message:

[W::vcf_parse_info] INFO/END=61067865 is smaller than POS at chr7:61067874

[friend1ws]

Thank you very much. These variants should be filtered due to the Too_small_size tags. However, I would like to develop a more general solution. These might be categorized as DUP or DUP:TANDEM instead of INS.

[RenzoTale88]

@friend1ws I've come across this too on a number of sites with v0.7.1. Several of them are not filtered for small size either.

[friend1ws]

OK. will try to solve in the next version.