Update CNV WDLs to WDL 1.0.

[samuelklee]

No description provided.

[ldgauthier]

Are these NUM_INTERVALS=${dollar}(grep -v '@' ${interval_list} | wc -l) shenanigans in ScatterIntervals solved in WDL 1.0?

[samuelklee]

If you're talking about the need for dollar, I'm not sure---looks like there are other instances of this hack in other WDLs, some of which have purportedly been updated to 1.0.

I don't think we would need this hack (in the gCNV WDLs, at least) if it weren't for the bash-related shenanigans in that task, which are required due to the wonky output style of IntervalListTools, though...

[ldgauthier]

We can certainly modify IntervalListTools to make the behavior more intuitive (e.g. add a new mode for INTERVAL_COUNT_WITH_OVREFLOW), but I'm not sure I understand the issue. We're just trying to avoid calling the GATK command if the scatter is going to be a noop?

The GATK SplitIntervals has slightly different behavior that's helpful in some cases.

[ldgauthier]

From gatk-workflows/gatk4-germline-cnvs#2
PreprocessIntervals doesn't need a sequence dictionary arg if it has a reference fasta

[samuelklee]

We already added the functionality needed for the gCNV workflow to IntervalListTools in broadinstitute/picard#1208. The issue is that the tool outputs each scattered interval list to a separate directory if the number of scatters is greater than 1, but it just outputs to a file (essentially a noop) if we don't need to scatter. Not sure the reason for this design, but it makes things difficult from the perspective of WDL. Would be easier if the expected output was always Array[File]+.

I don't really see why IntervalListTools needs to create those intermediate directories (nor why the naming scheme is determined by DecimalFormat("0000")---don't think this is documented, either), but I am not sure if that behavior is expected by now or if it is safe to modify it.

Pretty sure SplitIntervals is just calling the same backend class used by IntervalListTools. Perhaps that tool might've been spun off before we exposed the Picard tools? See e.g. #5392 (comment). I think we should try to avoid writing such custom/utility GATK tools unless really warranted.

[samuelklee]

Actually note broadinstitute/picard#1208 (comment) concerning the naming scheme of the output directories.

[mwalker174]

Resolved via #6506