Nextflow: samplesheet rows to Galaxy collections
Use this when the source evidence is a samplesheet or manifest reader that emits repeated channel records such as tuple(meta, path(reads)).
Source idiom
val(meta)carries identity and attributes; it is not itself a Galaxy dataset.- Repeated
tuple(meta, path(file))usually maps to a Galaxylistcollection. - Repeated
tuple(meta, [R1, R2])usually maps tolist:paired. - One standalone pair may be
paired, but repeated samples should staylist:paired.
Galaxy target choice
| Source row payload | Galaxy shape | Implementation route |
|---|---|---|
| One file per row | list | tabular-to-collection-by-row or direct collection input |
| Forward/reverse reads per row | list:paired | collection-build-list-paired-with-apply-rules |
| One pair total | paired | Direct paired input or paired collection construction |
| Mixed single/paired rows | split collections or review | source review trigger |
| Extra non-read files | parallel collections or review | source review trigger |
Implementation route
- Preserve stable sample identity from the row metadata.
- Decide collection type from file payload shape, not from metadata object shape alone.
- If the source starts as a tabular manifest, use tabular-to-collection-by-row.
- If pairedness is encoded in columns or identifiers, use collection-build-list-paired-with-apply-rules.
- If the workflow maps domain tools per row, use manifest-to-mapped-collection-lifecycle.
- After mapped steps, relabel, reshape, or sync only as needed through galaxy-collection-patterns.
Evidence posture
This page is grounded in existing Foundry research notes, especially nextflow-to-galaxy-channel-shape-mapping and iwc-map-over-lifecycle-survey. Generated Nextflow fixtures were not present during authoring, so do not treat this page as direct fixture evidence.