chore: clean up Nextflow selector warnings and dead config

[m-huertasp]

What this does

Removes accumulated Nextflow "no matching selector" warnings that appeared after recent feature additions deleted or renamed processes. The fix has two parts: stripping dead withName blocks for processes that no longer exist, and simplifying overly-qualified selector paths so they match the actual module names. Feature-gated processes (regressions, introns, non-prot) get explicit comments so the expected warning is documented rather than treated as a regression.

Changes

• Remove dead withName selectors for deleted processes: CREATESAMPLEPANELS, SUBSETPANEL, MUTRATE, SUBSETPILEUP, READSPOSBED
• Simplify overly-qualified selector paths in base.config, exome.config, modules.config, panels.config (drop redundant namespace prefixes that no longer match)
• Pin singleton channels at source in workflows/deepcsa.nf and subworkflows; remove redundant .first() calls

What to review

Run in Seqera to make sure no errors are raised there.

There are still some warnings from selectors:

WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:.*INTRONS:.*
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:.*PROT:.*
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:.*NONPROT:.*
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:.*EXONS:.*
WARN: There's no process matching config selector: .*PLOTDEPTHSALLCONS:DEPTHSSUMMARY
WARN: There's no process matching config selector: .*PLOTDEPTHSEXONS:DEPTHSSUMMARY
WARN: There's no process matching config selector: .*NONPROT:SUBSETMUTPROFILE
WARN: There's no process matching config selector: .*MULTI:SUBSETOMEGA
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGAMULTI:.*
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGANONPROTMULTI:.*
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGA:.*GLOBALLOC
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGAMULTI:.*GLOBALLOC
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGANONPROT:.*GLOBALLOC
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGANONPROTMULTI:.*GLOBALLOC
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:MUTDENSITYPROT:SUBSETMUTDENSITY
WARN: There's no process matching config selector: .*NONPROT:SUBSETMUTDENSITY
WARN: There's no process matching config selector: .*SYNONYMOUS:SUBSETMUTDENSITY
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:EXPECTEDMUTATEDCELLS:.*
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGA:SITECOMPARISON
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGA:SITECOMPARISONMULTI
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGA:SITECOMPARISONGLOBALLOC
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGA:SITECOMPARISONGLOBALLOCMULTI
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGAMULTI:SITECOMPARISON
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGAMULTI:SITECOMPARISONMULTI
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGAMULTI:SITECOMPARISONGLOBALLOC
WARN: There's no process matching config selector: BBGTOOLS:DEEPCSA:OMEGAMULTI:SITECOMPARISONGLOBALLOCMULTI

What I did to remove most of these "selector" warnings was to follow what is done in here. Basically, for those selectors that can be distinguished on their own, use only their name:

• Only an explicit module name can cope with 'optional' execution and have a selector still picked up, even if it's 'turned off'.
• Fully resolved paths or wildcarded withName selectors cannot be evaluated by Nextflow in this manner, and thus give you warnings like above.
  For these selectors that I couldn't use the explicit module name, it is because some logic is used to use specific subfolders based on the name (e.g. .INTRONS., etc) or because same module name is used (e.g. BBGTOOLS:DEEPCSA:OMEGAMULTI:SITECOMPARISON and BBGTOOLS:DEEPCSA:OMEGA:SITECOMPARISON

For the .first() redundant calls, I think I fix most of them. There are some missing related with the clonal_structure and the get_signatures modes that I couldn't find.

Testing

Verified by running the pipeline. No functional behaviour change found.

closes #468