dellytools / delly

DELLY2: Structural variant discovery by integrated paired-end and split-read analysis
BSD 3-Clause "New" or "Revised" License
413 stars 136 forks source link

RDCN tag in Delly vcf output #371

Open themkdemiiir opened 4 months ago

themkdemiiir commented 4 months ago

Hello,

Thank you for providing such great tools. I appreciate it. I have a few questions about the output vcf. I ran the HG002 sample with default parameters for germline call, but I noticed that the RDCN values for DEL, INV, and DUP structural variant types were varying. I need clarification about the meaning of this key since it could also be used to indicate genotype. Please clarify this for me.

Also, another question: how do I filter my WGS sample if it's not cohort? Should I remove lines that are imprecise or don't pass filters?

tobiasrausch commented 4 months ago

QUAL and FILTER are good fields to filter on, as well as PE, MAPQ or SR, SRMAPQ for IMPRECISE and PRECISE variants, respectively. Delly is primarily an SV breakpoint caller but if you only want simple deletions that do show a read-depth decrease or simple duplications that do show a read-depth increase then RDCN is useful for filtering.

themkdemiiir commented 4 months ago

QUAL and FILTER are good fields to filter on, as well as PE, MAPQ or SR, SRMAPQ for IMPRECISE and PRECISE variants, respectively. Delly is primarily an SV breakpoint caller, but if you only want simple deletions that show a read-depth decrease or simple duplications that do show a read-depth increase, then RDCN is useful for filtering.

Thank you for your answer. I am having difficulty identifying possible fusions because the breakpoint mate's orientation is not shown. Can the breakpoint vcf line be split into two? Also, I am receiving RDCN for inversion. What does this mean?

Other questions:

tobiasrausch commented 4 months ago

Delly uses INFO/CT to define the breakpoint orientations, this is explained at length in various delly-users group posts, e.g.: delly-users

themkdemiiir commented 4 months ago

Please show the CT orientations that match with the following pairs: W-Y (5 to 5), V-U (5 to 3), and X-Z (3 to 3). I assume these match. What do you think?

REF ALT Meaning
s t[p[ piece extending to the right of p is joined after t
s t]p] reverse comp piece extending left of p is joined after t
s ]p]t piece extending to the left of p is joined before t
s [p[t reverse comp piece extending right of p is joined before t
The example in Figure 1 shows a 3-break operation involving 6 breakends. It exemplifies all possible orientations
of breakends in adjacencies. Notice how the ALT field expresses the orientation of the breakends.
#CHROM POS ID REF ALT QUAL FILTER INFO
2 321681 bnd W G G]17:198982] 6 PASS SVTYPE=BND
2 321682 bnd V T ]13:123456]T 6 PASS SVTYPE=BND
13 123456 bnd U C C[2:321682[ 6 PASS SVTYPE=BND
13 123457 bnd X A [17:198983[A 6 PASS SVTYPE=BND
17 198982 bnd Y A A]2:321681] 6 PASS SVTYPE=BND
17 198983 bnd Z C [13:123457[C 6 PASS SVTYPE=BND