Auxiliary modules¶

These workflows are run separately from Modules 0-7. samwise_dir identifies the SAMWISE source tree; working_dir identifies the results root. If either is omitted, samwise_dir defaults to the launched workflow's directory and working_dir defaults to samwise_dir.

Auxiliary Module 1: Assembly annotation¶

AuxModule_1_assemblyAnnotate.nf functionally annotates assemblies, not MAGs. It has no Module 1 read-trimming prerequisite. It discovers assembly directories produced by Module 2, Module 2B, and Module 5, retains scaffolds at least min_scaffold_bp bases long, removes exact and reverse-complement-equivalent full-length duplicates, combines the retained scaffolds, and runs EggNOG-mapper.

nextflow run AuxModule_1_assemblyAnnotate.nf \
  --working_dir ./output_samwise \
  --threads 36 \
  --min_scaffold_bp 1000

At least one of these directories must exist:

• <working_dir>/module_2_readassembly/assemblies
• <working_dir>/module_2b_coassembly/assemblies
• <working_dir>/module_5_subassembly/assemblies

The workflow fails if none exists or if min_scaffold_bp is less than 1. Missing optional source directories are skipped. Managed published directories are cleaned before a new run.

Parameters¶

Database and outputs¶

A valid database at <working_dir>/module_6_magannotate/databases/eggnog is reused before another database is downloaded. An explicit eggnog_data_path or eggnog_data_dir takes precedence; otherwise the auxiliary database is under <outdir>/databases/eggnog. diamond requires a valid DIAMOND database; mmseqs requires a valid MMseqs database. Unsupported methods and missing databases fail validation or execution.

The output contains:

• filtered_assemblies/: retained per-source assembly FASTA files.
• inputs/eggnog_assembly_scaffolds.fasta: combined EggNOG input.
• eggnog/: EggNOG-mapper output files.
• summary/assembly_filtering_manifest.tsv, scaffold_filtering_manifest.tsv, and scaffold_filtering_stats.tsv.
• summary/eggnog_scaffold_manifest.tsv and eggnog_status.tsv.
• summary/assembly_annotation_run_summary.tsv and cleanup status.
• logs/: preparation and EggNOG logs; setup/ contains setup status.

Auxiliary Module 2: MVP viral analysis¶

AuxModule_2_mvp.nf prepares compatible inputs from Module 2 individual assemblies, optional Module 2B coassemblies, and optional Module 5 subtractive assemblies, then invokes selected MVP stages.

nextflow run AuxModule_2_mvp.nf \
  --working_dir ./output_samwise \
  --mvp_modules "0,1,2,3,4,5,100" \
  --threads 36 \
  --memory_gb 0

Inputs and manifests¶

By default, the workflow reads the following paths under working_dir:

• Module 2 assemblies: <working_dir>/module_2_readassembly/summary/assembly_manifest.tsv.
• Module 1 trimmed reads: <working_dir>/module_1_readtrimming/summary/trimmed_manifest.tsv.
• Module 2B assemblies: <working_dir>/module_2b_coassembly/summary/assembly_manifest.tsv.
• Module 2B reads: <working_dir>/module_2b_coassembly/summary/coassembly_trimmed_manifest.tsv.
• Module 5 assemblies: <working_dir>/module_5_subassembly/summary/subtractive_assembly_summary_manifest.tsv.
• Module 5 assembly files: <working_dir>/module_5_subassembly/assemblies.
• Module 5 unmapped reads: <working_dir>/module_5_subassembly/unmapped_reads.

Override these paths with assembly_manifest, trimmed_manifest, coassembly_manifest, coassembly_trimmed_manifest, subtractive_assembly_manifest, subtractive_assembly_dir, and subtractive_unmapped_reads_dir. Individual, coassembly, and subtractive inputs are enabled by their corresponding inclusion flags. At least one class must be enabled. With the default include_individual_assemblies=true, the Module 2 assembly manifest and Module 1 trimmed-read manifest must exist and contain matching, non-empty files. Missing or empty Module 2B and Module 5 manifests are skipped safely even when their inclusion flags remain true; records that are present must still resolve to non-empty assembly and read files.

Paired reads are interleaved record-by-record with BBTools only when required; existing interleaved reads are used directly. The prepared read manifest and metadata are generated before MVP runs.

Parameters¶

Validation, ordering, and outputs¶

Valid stage values are 0,1,2,3,4,5,6,7,99,100. Empty or invalid selections fail. Selected stages always run in this canonical order, regardless of the order supplied: 0 (setup), 1 (geNomad/CheckV), 2 (filter), 3 (clustering), 4 (read mapping), 5 (vOTU table), 6 (functional annotation), 7 (binning), 99 (MIUViG), then 100 (summary). Omitting an earlier stage means its required outputs must already exist under <outdir> from a previous run.

The workflow validates read_type, normalization, and filtration, rejects conflicting geNomad modes, requires MIUViG settings when stage 99 is selected, and requires a template for prep_submission. MVP and BBTools environments are installed or reused and validated; disabling auto-install requires valid existing environments. Database installation is disabled by default. MVP uses <outdir>/00_DATABASES as the database root unless mvp_database_dir is supplied. genomad_db_path and checkv_db_path override the corresponding databases independently. When install_databases=false, the required databases must already exist; when it is true, MVP may install missing databases under the selected database root.

Outputs are written directly to <outdir> so completed MVP outputs remain available if a later stage fails. Important managed outputs include:

• interleaved_reads/: generated interleaved FASTQ files when needed.
• metadata/mvp_prepared_reads_manifest.tsv and metadata/mvp_metadata.tsv.
• summary/mvp_read_preparation_stats.tsv, input summaries, and completion status.
• logs/: setup, preparation, command, and run logs.
• 00_DATABASES/: MVP databases when installed there.
• MVP's stage directories and files, using the names expected by MVP.