search

arq5x / lumpy-sv

lumpy: a general probabilistic framework for structural variant discovery
MIT License
307 stars 119 forks source link

Lumpy specifies an inversion with two breakends only #325

Closed jazberna1 closed 4 years ago

jazberna1 commented 4 years ago

Hello,

I have run a tumour/nomal bam pair through Lumpy. There is one particular SV that Lumpy finds which could be a potential inversion as another tool (Delly) finds it too. However Lumpy does not produce the four records (breakends) that specify an inversion according to the vcf specifications (https://samtools.github.io/hts-specs/VCFv4.3.pdf). Instead, I only get two records in the vcf for such inversion. Is there a reason for that? I have pasted below the Lumy output for that variant as well as the Delly output.

Many thanks Jorge

Lumpy: two records instead of four for an inversion 16 52043343 3550_1 N [16:52052737[N 23.68 . SVTYPE=BND;STRANDS=--:7;IMPRECISE;CIPOS=-385,29;CIEND=-563,29;CIPOS95=-81,8;CIEND95=-140,5;MATEID=3550_2;E VENT=3550;SU=7;PE=7;SR=0;AC=1;AN=2 GT:GQ:SQ:GL:DP:RO:AO:QR:QA:RS:AS:ASC:RP:AP:AB 0/1:23:23.68:-11,-8,-39:53:46:6:46:6:0:0:0:46:6:0.12 16 52052737 3550_2 N [16:52043343[N 23.68 . SVTYPE=BND;STRANDS=--:7;IMPRECISE;CIPOS=-563,29;CIEND=-385,29;CIPOS95=-140,5;CIEND95=-81,8;MATEID=3550_1;E VENT=3550;SECONDARY;SU=7;PE=7;SR=0;AC=1;AN=2 GT:GQ:SQ:GL:DP:RO:AO:QR:QA:RS:AS:ASC:RP:AP:AB 0/1:23:23.68:-11,-8,-39:53:46:6:46:6:0:0:0:46:6:0.12

Delly: one record that specify and inversion 16 52043343 INV00008128 T <INV> . PASS IMPRECISE;SVTYPE=INV;SVMETHOD=EMBL.DELLYv0.7.9;CHR2=16;END=52052737;PE=7;MAPQ=43;CT=5to5;CIPOS=-346,346;CIEND=-346,346 GT:GL:GQ:FT:RCL:RC:RCR:CN:DR:DV:RR:RV 0/1:-16.7753,0,-221.751:10000:PASS:773:1365:704:2:42:7:0:0

ryanlayer commented 4 years ago

Wow. The SV section of the VCF spec has been grant extended since we added our VCF output module. We do not follow this version of the spec, which would explain the difference.

That said, I would be a bit suspicious of this INV because only negative strand pairs were observed (look at LUMPY’s STRAND field). The — and ++ pairs should roughly equal.

On Jan 19, 2020, at 1:52 AM, Jorge Zamora notifications@github.com wrote:

 Hello,

I have run a tumour/nomal bam pair through Lumpy. There is one particular SV that Lumpy finds which could be a potential inversion as another tool (Delly) finds it too. However Lumpy does not produce the four records (breakends) that specify an inversion according to the vcf specifications (https://samtools.github.io/hts-specs/VCFv4.3.pdf). Instead, I only get two records in the vcf for such inversion. Is there a reason for that? I have pasted below the Lumy output for that variant as well as the Delly output.

Many thanks Jorge

Lumpy: two records instead of four for an inversion 16 52043343 3550_1 N [16:52052737[N 23.68 . SVTYPE=BND;STRANDS=--:7;IMPRECISE;CIPOS=-385,29;CIEND=-563,29;CIPOS95=-81,8;CIEND95=-140,5;MATEID=3550_2;E VENT=3550;SU=7;PE=7;SR=0;AC=1;AN=2 GT:GQ:SQ:GL:DP:RO:AO:QR:QA:RS:AS:ASC:RP:AP:AB 0/1:23:23.68:-11,-8,-39:53:46:6:46:6:0:0:0:46:6:0.12 16 52052737 3550_2 N [16:52043343[N 23.68 . SVTYPE=BND;STRANDS=--:7;IMPRECISE;CIPOS=-563,29;CIEND=-385,29;CIPOS95=-140,5;CIEND95=-81,8;MATEID=3550_1;E VENT=3550;SECONDARY;SU=7;PE=7;SR=0;AC=1;AN=2 GT:GQ:SQ:GL:DP:RO:AO:QR:QA:RS:AS:ASC:RP:AP:AB 0/1:23:23.68:-11,-8,-39:53:46:6:46:6:0:0:0:46:6:0.12

Delly: one record that specify and inversion 16 52043343 INV00008128 T . PASS IMPRECISE;SVTYPE=INV;SVMETHOD=EMBL.DELLYv0.7.9;CHR2=16;END=52052737;PE=7;MAPQ=43;CT=5to5;CIPOS=-346,346;CIEND=-346,346 GT:GL:GQ:FT:RCL:RC:RCR:CN:DR:DV:RR:RV 0/1:-16.7753,0,-221.751:10000:PASS:773:1365:704:2:42:7:0:0

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

jazberna1 commented 4 years ago

Hi,

Thanks so much! That was very clarifying.

Jorge