CNVs are not gentyped

[Damian Loska]

Hi,

I've done such steps:

SV part:
1) For each sample separately:
delly call -g ref.fa -x human.hg38.excl.tsv pat1.bam -o dpreSV.pat1.bcf

2) Togehter for all *bcf from point above
delly merge -o sitesSV.bcf /delly/preSV/dpreSV*bcf

CNV part:
serveral sapearted samples:
delly cnv -o pCNV_pat1.bcf -g ref.fa -m delly/MAP/map.fa -l sitesSV.bcf pat1.bam 

merge all BCFs from step above:
delly merge -e -p -o sites_cnv_1k_100k.bcf -m 1000 -n 100000 delly/preCNV/pCNV_*bcf

delly cnv -u -v sites_cnv_1k_100k.bcf -g ref.fa -m delly/MAP/map.fa -o cnv_pat1.bcf pat1.bam


but in the output cnv_pat1.bcf has no genotypes. All looks like ./.
 GT:CN:CNL:GQ:FT:RDCN:RDSD       ./.:0:0,-1001.2,-1001.2,-1001.2,-1001.2,-1001.2,-1001.2,-1001.2,-1001.2,-1001.2:10012:PASS:0:0.025

How to obtain GT information?

[Damian Loska]

At least the samples with 0/0 could be marked. Now I don't know which sample has the CNV and which not...

--
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...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/delly-users/026dd21f-5ba3-479d-82fd-cc4a21c5ddb1n%40googlegroups.com.

[Víctor Fernández]

Hi there,

I have exactly the same problem. In my case, even for separate calls, the CNVs appear as missing genotype (./.).

Regarding your second message, you could access to the information of each sample in the merged file. Then, if the CN field is 0, you can consider that sample as 0/0 because it does not present the CNV (that's what I undestand).

I post my message here so that if anyone knows how to solve this, I can also see the answer. 

[Víctor Fernández]

Hi there, 

I have exactly the same problem. In my case, even for separate calls, the CNVs appear as missing-genotype (./.). 

Regarding your second message, you could access to the information of each sample in the merged file. Then, if the CN field is 0 you can consider that sample as 0/0, because it does not present the CNV (that's what I undestand). 

I put my message here so if anyone knows how to solve this I can also see the answer. 

El viernes, 30 de agosto de 2024 a las 12:23:27 UTC+2, Damian Loska escribió:

[tr]

Hi,

I can confirm that for CNV calling delly only uses the FORMAT:CN field at the moment. That's because of multi-allelic CNVs in the population where you might have a range of CN values. One could encode this as multiple ALTs but that's not the case at the moment.

Best, Tobias