Refactor joint germline calling

[maxulysse]

So this established back the old behavior. The pipeline was not supposed to do filtering for joint germline, but single sample only, (where it sometimes can be skipped because it takes a long time): https://gatk.broadinstitute.org/hc/en-us/articles/360035535932-Germline-short-variant-discovery-SNPs-Indels-

I think we might need to refactor this subworkflow a bit to make it more intuitive. and put the whole FilterVarianttranches/CNNScore part into a singlesample haplotypecaller subworkflow, but I would do that in a separate PR

Originally posted by @FriederikeHanssen in #1050 (comment)