Create md5 sums for data upload to public repositories

Public data repositories such as GEO require md5 sums to be included in the metasheet containing the experimental details. This can be quite a painful process to carry out down the line but it's something that can be automated during the pipeline. I'd imagine that the md5 sums would be generated for the raw fastqs, processed bam, peaks and bigwig files. Would also have to parse the Picard insert size metrics file to get the insert mean and std dev. These could all be collected at the end of the pipeline and placed in a tsv file that could then be copy and pasted appropriately into the metasheet.

Start by reviewing where the pipeline produces the raw FASTQs, processed BAMs, peaks, bigWigs, and Picard insert-size metrics. Define an end-of-pipeline TSV containing MD5 sums for those outputs plus the insert mean and standard deviation, suitable for copying into a repository metasheet. No specific files or tests are named, so the pipeline entry point and validation approach must be located first.