Andrew Birnberg

Thanks for looking into this and coming up with a solution! I'm happy to run my dataset through and report back. It looks like the PR is still building in...

Perfect. Just ran it and the result is as advertised. I used the resulting annotated bam (following sorting and addition of read groups) with `picard UmiAwareMarkDuplicatesWithMateCigar` and encountered no issues....

No issues detected after running through the suggested pipeline of additional tools.

While I have your attention, I'd like to ask about the use case for `GroupReadsWithUmi` as opposed to a tool like `UmiAwareMarkDuplicatesWithMateCigar`. It seems that when you're using UMI's it's...

Got it. My use case is different but it makes sense to do more aggressive quality filtering for variant detection. Thank you for your help on this!

Just figured out that building and serving the frontend from MacOS (still from Docker) works, but not from a Centos 7 instance. Not sure why that would be the case...

Great, thanks for that tip! As you say, it works if you leave off the `val` qualifier.

My preference would be to add an optional argument to `publishDir`, such as `contentType` or `mimeType`, which is only used by storage plugins, such as S3 or GCP, for which...

I agree, it would be nice to have a new process directive where you could provide a filename (or pattern/extension) to content-type map that can be configured globally or for...

+1 for this, however, I'd be happy with at least allowing the default `$ref` implementation in `jsonschema`, which allows schema-internal references without additional configuration. Not supporting `$ref` for internal schema...