Getting same winner HLAs everytime

[emrahselli]

I get the same alleles hla a_01_01_01, hla b_07_02_01 and hla c_01_02_01 when I run WES bam files from different individuals. When I run the test.bam, only hla c second winner did not match with the original winners. Any help would be greatly appreciated. Thanks

[jason-weirather]

Hi @emrahselli , this is not an issue I've run into myself either through this fork or through the proper polysolver software. If I were to guess at the problem, I wonder if this could be a coverage issue (shallow sequencing), or even an alignment reference genome issue (this fork is only compatible with hg19), but I am probably not best suited to debug this problem. I recommend you try the official polysolver docker https://hub.docker.com/r/sachet/polysolver/tags/ and if you still run into a problem to contact the author Sachet Shukla. He also includes some test data for you to run, and you could give that a try too. If you are unable to get the test data to work, its most likely an issue with your install or environment.

(edit) Just saw your notes on running the test. I think your result is the same I had with the test. (this fork is not a perfect match for whatever the original environment and reference for polysolver was).

[emrahselli]

Thank you very much @jason-weirather . You may be right about coverage issue. I will try the original environment to see anything will change. Thanks again!

[wentgithub]

do you know whether POLYSOLVER can be used for small panel data?
thanks a lot

[jason-weirather]

do you know whether POLYSOLVER can be used for small panel data?
thanks a lot

Hi @wentgithub , that's probably better directed at the polysolver authors. I would speculate that it will be fine if the class I genes of the various haplotypes are included in the capture of your small panel since I think polysolver only crunches on that portion of the genome (not 100% sure about that). I've never tested that myself or know of anyone running it that way.

[wentgithub]

@jason-weirather
Thanks a lot

[Nitin123-4]

Make sure you are using /home/polysolver/scripts/shell_call_hla_type not /home/polysolver/scripts/shell_call_hla_type_test.

Also, check there should not be any error after the run is over.