using gatk haplotypecaller for variants extraction

Question

[Nemo]

Hi,

I have rna-sequenced data from covid patients. I am using hisat2 for aligning the reads to reference. So, the resulted bam files after indexing are now ready. I would like to use gatk happlotypecaller for extracting variants from my bam files.

First, before calling haplotypecaller should I do any further analysis on my bam files?

Second, in the gatk forum HaplotypeCaller I see the sample commands but I noticed they are for DNA seq data. So, it means I cannot use gatk haplotypecaller for rna seq data? If yes, can you give me an example what would be the value for -ERC flag? if no, what else should I use?

Thanks in advance for any help.

[Ram]

I see the sample commands but I noticed they are for DNA seq data. So, it means I cannot use gatk haplotypecaller for rna seq data?

You can use HC for RNAseq data.

Did you try Googling your question? I Googled "variant calling from rnaseq gatk" and the first result is this: https://gatk.broadinstitute.org/hc/en-us/articles/360035531192-RNAseq-short-variant-discovery-SNPs-Indels-

[Nemo]

thanks @Ram, yes I searched for my question, and I read the link you sent. As I found I can use gatk haplotypecaller (I am skeptical yet) but I dont know what option should I pick for -ERC flag, and what is the differences? Can you confirm that I can use gatk haplotypecaller for my rna seq data?, and guide me on choosing the right option for -ERC flag?

Also, can you please send me a link to the HC tool you suggested?

[Ram]

I should have been clearer - I used HC as short for HaplotypeCaller. I'm used to using that because when I started off, pipelines were using UnifiedGenotyper and I was comparing the two. Instead of typing those every time, I would say UG vs HC. I think that short form is used widely as well, since Googling "gatk hc" brings up a ton of HaplotypeCaller articles.

Anyway, there is a best practices article for GATK + Variant Calling + RNAseq data. That might help although from what I recall, GATK variant calling using RNAseq data is not well developed compared to using WGS/WES data. I recall using samtools/bcftools to call variants on RNAseq BAMs to check for concordance with MuTect2 and being sorely disappointed at the lack of concordance. You may not want to do variant calling with RNAseq.