Are there Best Practices for calling variants in RNAseq data?
We are working on updating our recommended workflow for calling variants in RNAseq data. Once that work is done and has been fully validated, we will publish detailed documentation explaining how it works and add it to our Best Practices documentation. In the meantime, you can find a work-in-progress version of the workflow on Github here. Note that the workflow is provided as-is wth no guarantees of either performance or correctness. The current workflow uses a combination of GATK 3.5 and GATK 4 beta versions.
Note also that we have not yet validated the germline short variants joint genotyping methods (HaplotypeCaller in -ERC GVCF mode per-sample then GenotypeGVCFs per-cohort) on RNAseq data. However, we are aware that some people have been trying out the joint genotyping workflow on RNAseq data and have not reported any major technical problems. You are welcome to try it on your own data, with the caveat that we cannot guarantee correctness of results, and will not be able to help you if something goes wrong. Please be sure to examine your results carefully and critically.