Hello guys, I work in a small lab, where we are using GATK best practices (3.8) - I have prepared the whole pipeline with bash scripts - like three steps - alignment, variant calling and annotation. We are using it mostly to analyse new WES data - like one set of 16 samples per month.
Now I would like to convert everything to a newer version of GATK 4 (and also create separate pipeline for hg38), I wonder which way is best for us. We have only local workstations, where we perform whole analysis - fastq to vcf for all samples.
Which way would you recommend the most? Rewrite bash scripts with newer .jar version of all tools, use prepared WDL scripts available on gatk website or using docker? (I have very little experience with docker container, but I am thinking that it could be a great opportunity to combine learning docker and preparing new pipelines in GATK4)
I am currently in the preparation and gathering the information, so all experiences and advices are very appreciated!
