Multi-Sample Output Organization

[apeltzer]

Is your feature request related to a problem? Please describe.

Currently, output is organized in a way that every VCF file lands in a single folder.

Describe the solution you'd like

It would be nice to have the possibility, if users can specify a single TSV containing multiple patients and the generated VCFs etc are stored nicely in

Describe alternatives you've considered

Creating n TSV files for each patient individually, which is quite cumbersome...

We discussed this together with @ggabernet @maxulysse on Gitter :-)

[ggabernet]

Currently it is possible to add as an input a TSV containing multiple patients, but I realized that somatic variant calling then calls all possibilities of tumor-normal pairs, even between different patients, so this would need to be fixed too.

[maxulysse]

I can see why this is happening, thanks for telling me about it

[ggabernet]

sure, I'm very happy to use your pipeline!

[maxulysse]

I'm glad that you're using it, with more user like you, Sarek can only improve its quality ;-)

[jongtaek-kim]

somatic variant calling then calls all possibilities of tumor-normal pairs, even between different
patients,

Has this been addressed yet? This wastes the time and computing resources exponentially as the number of patients increase.
I am resorting to running each patient T/N separately.

[maxulysse]

Hi,
I'm sorry, I didn't manage to find time to implement that yet.
But it's definitively something that I plan for our next release.

[maxulysse]

The bug encountered by @ggabernet will be fixed by #728.
Concerning the organization of files, is this just the vcfs that are posing an issue for you?
or would you want separate directories for the bams as well?

[ggabernet]

I would say the vcfs would be enough. Thanks for checking this!

…

On 14. Feb 2019, at 16:56, Maxime Garcia ***@***.***> wrote: The bug encountered by @ggabernet <https://github.com/ggabernet> will be fixed by #728 <#728>. Concerning the organization of files, is this just the vcfs that are posing an issue for you? or would you want separate directories for the bams as well? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#691 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ADyhf8ixgEvFvhfTrEMd4RtLLBlikZlSks5vNYc8gaJpZM4YmZu->.

[maxulysse]

Thanks for your input.
@jongtaek-kim what is your use case?

[jongtaek-kim]

@maxulysse
I am running the pipeline on exome T/N samples to detect somatic snvs and indels in my academic health center's HPC. Thanks for the pipeline repo.

[maxulysse]

@jongtaek-kim I can see why you're interested in this issue.
I does indeed makes a lots of sense to have that for exome.

Are VCFs organized in patient specific folder enough for you ?
Are you interested in the same thing for BAMs as well ?

[jongtaek-kim]

Are VCFs organized in patient specific folder enough for you ?
Are you interested in the same thing for BAMs as well ?

After doing some thinking, I think both are good ideas to implement.

[maxulysse]

OK, so the VCFs are to be separated in PR #728
I'll work on the Bams later on

[jongtaek-kim]

@maxulysse Thank you so much.

Also, appreciate getting the targetBED fixed.
eb3dfb1
Did little test run on dev branch and can attest --targetBED parameter worked.
This goes well with my exome sequencing where I have to provide a bed file.
Will run the whole thing with #728 merged dev and will see how it goes.