Hi all, Second question of the day, I am sorry.
I have been facing an issue that I saw many times on the forum but each time it seemed that you fixed it with new versions. Problem is I am using the very last version of GATK = 3.4 (downloaded yesterday).
I am using HaplotypeCaller single sample mode with all parameters by default as indicated in the "DNASeq" example in the HaplotypeCaller documentation page.
I have a problem with one variant rs206076 on BRCA2 chromosome 13 position 32915005. I have Sanger validation for this variant in all my samples and Unified Genotyper was calling it correctly.
Using HaplotypeCaller, there are some variants where the HOM call is made correctly and with very high quality (22 000) and others where HC calls the position as heterozygote with very low quality when it is obvious from IGV that the position is homozygote.
I put in the IGV screenshot on individual that fails (up) and the other that is ok (down).
I do have the same problem with another variant close to this one (not shown) whereas a couple of variants in between are called correctly in all samples.
Any idea about this issue ? thanks a lot Manon
I am finding that HaplotypeCaller is calling heterozygotes when there is no evidence when you look in IGV. Is this still a known problem?
I am using version 2.8-1-g932cd3a. Attached are two screen shots in IGV where a heterozygotes is wrongly called: