Closed Tintest closed 6 months ago
I am just another user, but for what it is worth I do not observe this oversegmentation when running the cnv module with a hg19 aligned DNA sample with 30kb bin sizes
Hello @TBradley27, thank you for your reply.
Have you done anything specific to your reference genome that might explain this difference, like having masked certain regions or something else ? As the BAM files were produced with wf-human-variation the only difference I suspect is at the genome level, but maybe I'm forgetting something. I'm personally experiencing the same problem with 30kb bins.
Thanks for your help, regards.
Hello @Tintest,
I was using the hs37d5 build of hg19, which uses the standard build along with some viral and decoy sequences (https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/phase2_reference_assembly_sequence/hs37d5.fa.gz)
It may be an idea to align your genome to this sequence or maybe another hg19 build to see if you see the same problems especially considering that the hg38 bin annotations are not canonical in the sense they were not released by QDNASeq authors
Hello @TBradley27, thank you for your reply.
I'll try with your genome build ASAP.
Thanks for your help, regards.
Hi @Tintest, we aren't aware of any specific build 38 issues, but please do let us know if @TBradley27's advice to try hg19 shows the same problem. You are correct, the --sex
parameter is not used by the CNV sub-workflow.
I've runed the pipeline with the hs37d5.fa.gz genome indicated by @TBradley27 but also with a version of the GRCh38 genome with some regions masked (including the PAR regions) and it hasn't changed anything, I still have the same problem. Do you have any ideas ?
Regards.
Hi @Tintest, sorry to hear that you're still experiencing difficulties. Currently, the only parameter that can be adjusted for QDNAseq is the bin size. Unfortunately, there are no additional modifications that we can make at this moment. However, it's worth mentioning that we are actively exploring alternative CNV callers, so I encourage you to stay tuned and keep an eye on our repository for any forthcoming updates.
Operating System
Other Linux (please specify below)
Other Linux
Debian GNU/Linux 11 (bullseye)
Workflow Version
v1.7.2
Workflow Execution
Command line
EPI2ME Version
No response
CLI command run
Workflow Execution - CLI Execution Profile
singularity
What happened?
Hello,
I'm writing to you because I've run the wf-human-variation CNV workflow on 3 PromethION sequenced samples. For all 3, whatever the size of the bin, I find myself with entire sex chromosomes called as CNVs. The number of segments is larger or smaller depending on the bin, while I have specified the sex (but I think it's not used by the CNV workflow). I haven't masked the PAR regions, but the problem seems to be more large. Sample 10 and 11 are male, sample 12 is female.
Is there anything I can do to avoid this? Is this a bug, or did I miss something ?
Regards.
Relevant log output
Application activity log entry
No response