We should dynamically configure some of the steps depending of the experiment size.

[ypriverol]

Description of feature

@daichengxin, @jpfeuffer:

When running multiple datasets, I have found that it would be good to have some logic to configure memory and CPU depending on the number of files. We have done that approach for the proteomicsLFQ step, however other processes could also benefit from it. Here is a list of process to review:

• NFCORE_QUANTMS:QUANTMS:TMT:PROTEININFERENCE:IDFILTER
• NFCORE_QUANTMS:QUANTMS:DIA:ASSEMBLE_EMPIRICAL_LIBRARY Issue quantms failing with a big dataset 15K files #339
• ID:DATABASESEARCHENGINES:SEARCHENGINESAGE

[daichengxin]

It should be possible to implement this in nextflow syntax. The logic can be divided two ways. What do you think?

1. number * unit usage.
2. range (24 G when number < 100, 128G when number<1000, 528G when number<10000, 1024G when number >10000)?
3. other?

[ypriverol]

Looks like a good strategy.