Delly filter outputs no variants in VCF

[PA]

Hi,

I genotyped one germline sample with DELLY for inversions using the command:

delly call -t INV -q 5 -u 5 -g <ref.fas> -o <out.bcf> <in.bam>

Apparently all ran OK and I got called variations in my VCF file. As an example, I got the following lines in the output:

scaffold25 1505074 INV00000933 A <INV> . PASS PRECISE;SVTYPE=INV;SVMETHOD=EMBL.DELLYv0.7.6;CHR2=scaffold25;END=1506348;PE=31;MAPQ=60;CT=3to3;CIPOS=-3,3;CIEND=-3,3;INSLEN=0;HOMLEN=2;SR=10;SRQ=0.990566;CONSENSUS=GGTTTGTCATCACAGGTTAGCTTTCGAGTTAAATATCCTCAAGATTATCATAAACAATTCTTGAACAACCAATTGACTACTCAAAACTAGTGCAAATAAGAGGTT;CE=1.91299 GT:GL:GQ:FT:RCL:RC:RCR:CN:DR:DV:RR:RV 0/1:-77.7521,0,-79.9513:10000:PASS:409:570:195:2:26:31:25:24

scaffold25 2144040 INV00000948 T <INV> . PASS PRECISE;SVTYPE=INV;SVMETHOD=EMBL.DELLYv0.7.6;CHR2=scaffold25;END=2160517;PE=16;MAPQ=54;CT=3to3;CIPOS=-11,11;CIEND=-11,11;INSLEN=0;HOMLEN=10;SR=10;SRQ=1;CONSENSUS=TTGTTCACTAGTAGTGATGTAAGCACGTCAATCTACATTTCCTTGATGTTATATCTCAACTGTAGAATGAAGAAATTAACACCCCATGGTCATCAATTCAGCGCTACCTGTA;CE=1.94853 GT:GL:GQ:FT:RCL:RC:RCR:CN:DR:DV:RR:RV 0/1:-61.8627,0,-39.5699:10000:PASS:4167:9741:4082:2:47:16:13:24

So, the filter is set to PASS and many reads support the SV (hence GT=0/1). Genotype Quality is 10000 (is this expected?).

But when I try to filter the VCF file with the DELLY filter command the resulting VCF is empty. I'm not sure how can this be as it looks like the SV calls seem fine to me. 

Maybe I'm missing something in interpreting DELLY output and any help would be great.

The command line for the filter was:

delly filter -p -f <in.bcf>

Many thanks,

Pedro

[Tobias Rausch]

Hi Pedro,

If you want to filter for somatic variants you need one tumor sample and at least one control as explained in the readme. To confidently identify germline variants a large sample cohort is required. "Large" really means here the more samples the better but at the very minimum something like 10 samples are required. If you only have a single sample then it is usually best to just filter the VCF manually for FILTER=PASS and INFO:PE (the higher the better) for imprecise variants and INFO:SR (the higher the better) for precise variants. The 2 inversions you posted look indeed fine because the paired-end and split-read support is very high.

Best, Tobias

--
You received this message because you are subscribed to the Google Groups "delly-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to delly-users+unsubscribe@googlegroups.com.
To post to this group, send email to delly...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[PA]

Hi Tobias,

that makes sense.

thanks for the help in interpreting DELLY results.

All the best.

Pedro

To unsubscribe from this group and stop receiving emails from it, send an email to delly-users...@googlegroups.com.