Dear all, thank you for your amazing work! I was looking at your pipeline and I was wondering if you have evaluated the possibility to integrate this pipeline as a workflow in nf-core/sarek. In this way a persona can choose to build the reference panel and automatically run sarek with one command. what do you think about it?
👍 1
m
Maxime Garcia
03/17/2025, 4:55 PM
no plans to do that
Maxime Garcia
03/17/2025, 4:56 PM
in my ideas, you'd need to run this pipeline on 40 normal samples to then run it on sarek and other downstream pipelines with the newly created panelreferences
👍 1
r
Rike Hanssen
03/25/2025, 1:58 PM
In fact, in general, we are moving away from integrating reference computation with pipelines. It seemed useful at first but turns out it is a big maintenance overhead, you typically would only want to run it once and then reuse the references over and over again and share the references between different pipelines (see #C073ZDLNC92)
m
Maxime Garcia
03/25/2025, 2:07 PM
and to be fair, pon aren't a different kind of references that the regular ones