search

Gabaldonlab / perSVade

perSVade: personalized Structural Variation detection
GNU General Public License v3.0
36 stars 5 forks source link

How can do joint calling for multiple samples #13

Closed DrChunLi closed 1 year ago

DrChunLi commented 1 year ago

Dear Sir: gridss suggest that "Joint calling should always be used for related samples", but I can NOT find any instruction for joint calling in the perSVade guide. How should I do to run joint calling?

MikiSchikora commented 1 year ago

Good morning,

We have currently not implemented a joint calling strategy, since perSVade is focused on providing variant calls for single samples. I am working on a module to compare the SVs found in different samples and keep sample-specific variants, which is conceptually similar to joint calling.

Imagine that you set of sample A (i.e. a tumor) variants and you want to remove the variants that are in another sample B (a matched normal sample). I'd use bedmap to define the variants from sample A that do not overlap any breakpoint from sample B. I'd use the unfiltered gridss results from sample B (provided in the perSVade output) to make sure that you remove all background variants.

I am working on this pipeline, but I can't guarantee when it is going to be tested and ready. Maybe you can implement such a pipeline that suits your samples?

Best, Miquel Àngel Schikora

MikiSchikora commented 1 year ago

Hi, Is this solved? I'd close the issue if so.

Miquel Àngel Schikora Barcelona Supercomputing Center

DrChunLi commented 1 year ago

Dear Sir: Thank you for your reply, the issue can be closed, I got the answer. Best wishes.

At 2023-02-02 19:10:22, "MikiSchikora" @.***> wrote:

Hi, Is this solved? I'd close the issue if so.

Miquel Àngel Schikora Barcelona Supercomputing Center

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>