nanopore.yaml

# Nanopore sequencing model configuration
# ONT Dorado basecalling + optional modkit methylation analysis

id: nanopore
name: Nanopore Sequencing
version: "1.0.0"
category: sequence_analysis
description: >
  Oxford Nanopore analysis workflow supporting POD5 re-basecalling via Dorado,
  direct tagged-BAM methylation summarization with modkit, and FASTQ plasmid QC.
container: dorado.sif

ui_icon: waveform
ui_color: "#0EA5E9"

workflow_family:
  id: ont_ngs
  analysis_owner: nextflow_analysis
  device_control_owner: bms_service_api
  manifest_contract: sequence_qc.manifest.v1
  canonical_workflows:
    - ont_basecall_dna
    - ont_basecall_rna
    - ont_plasmid_qc
    - ont_construct_screening
    - ont_methylation_analysis
    - ont_fastq_qc
    - ont_pooled_reference_assignment
    - wf_clone_validation

quality_modes:
  molecule_types:
    - dna
    - rna
  basecalling_modes:
    - simplex
    - duplex
  dorado_quality_modes:
    - fast
    - hac
    - sup
  default_quality_mode: sup
  default_basecalling_mode: simplex
  default_device: cuda:0
  device_examples:
    - cuda:0
    - cuda:all
    - cpu
  modified_base_models:
    - none
    - 5mC_5hmC
    - 6mA
  barcode_fields:
    - barcode_kit
    - sample_sheet

inputs:
  - pod5
  - bam
  - fastq
  - fasta
  - reference_set_manifest
outputs:
  - bam
  - bed
  - tsv
  - json
  - html
  - fasta

sequence_qc_contract:
  manifest: qc_manifest.json
  manifest_schema: sequence_qc.manifest.v1
  fastq_qc_artifacts:
    - fastq_qc_summary.tsv
    - fastq_alignment_stats.tsv
    - fastq_coverage.tsv
    - per_base_support.tsv
    - qc_manifest.json
    - fastq_consensus.fasta
    - fastq_consensus.fasta.fai
    - igv_track_config.json
    - igv_report.html

modes:
  - id: basecall_dna
    name: ONT DNA Basecalling
    description: Dorado DNA basecalling from existing POD5 run outputs. Legacy FAST5 must be converted and model-qualified outside this workflow.
    params:
      - pod5_dir
      - reference_fasta
      - dorado_model
      - dorado_device
      - min_qscore
      - trim_adapters
      - emit_summary
      - dorado_batch_size

  - id: basecall_rna
    name: ONT RNA Basecalling
    description: Dorado RNA basecalling from existing ONT run outputs.
    params:
      - pod5_dir
      - reference_fasta
      - dorado_model
      - dorado_device
      - min_qscore
      - trim_adapters
      - emit_summary
      - dorado_batch_size

  - id: plasmid_qc
    name: ONT Plasmid QC
    description: Reference-backed plasmid QC from exactly one POD5, BAM, or FASTQ input.
    params:
      - pod5_dir
      - bam_path
      - fastq_path
      - reference_fasta
      - dorado_model
      - dorado_device
      - min_qscore
      - trim_adapters
      - emit_summary
      - dorado_batch_size
      - bam_force_realign
      - bam_min_mapq
      - run_fastq_qc
      - expected_plasmid_size
      - samtools_consensus_config
      - min_fastq_read_length
      - fastq_minimap2_preset
      - fastq_minimap2_allow_secondary
      - igv_track_window_bp
      - igv_report_max_sites
      - igv_report_flanking_bp
      - enable_rotating_reference_frames
      - rotation_scan_step_bp
      - single_ref_split_min_mapq
      - single_ref_split_min_segment_bp
      - single_ref_split_max_query_gap_bp

  - id: construct_screening
    name: ONT Construct Screening
    description: Expected-construct screening from exactly one POD5, BAM, or FASTQ input.
    params:
      - pod5_dir
      - bam_path
      - fastq_path
      - reference_fasta
      - dorado_model
      - dorado_device
      - min_qscore
      - trim_adapters
      - emit_summary
      - dorado_batch_size
      - bam_force_realign
      - bam_min_mapq
      - run_fastq_qc
      - expected_plasmid_size
      - samtools_consensus_config
      - min_fastq_read_length
      - fastq_minimap2_preset
      - fastq_minimap2_allow_secondary
      - run_assembly
      - wf_clone_sample
      - wf_clone_large_construct
      - wf_clone_approx_size
      - wf_clone_assm_coverage
      - wf_clone_trim_length
      - wf_clone_min_quality
      - wf_clone_assembly_tool
      - wf_clone_basecaller_model
      - wf_clone_primers
      - wf_clone_insert_reference
      - wf_clone_host_reference
      - wf_clone_regions_bedfile
      - wf_clone_flye_quality
      - wf_clone_non_uniform_coverage
      - wf_clone_canu_fast
      - wf_clone_cutsite_mismatch
      - wf_clone_primer_mismatch
      - wf_clone_expected_coverage
      - wf_clone_expected_identity

  - id: fastq_qc
    name: ONT FASTQ QC
    description: FASTQ read QC plus reference-backed alignment/plasmid evidence when a reference is supplied.
    params:
      - fastq_path
      - reference_fasta
      - expected_plasmid_size
      - samtools_consensus_config
      - run_fastq_qc
      - min_fastq_read_length
      - fastq_minimap2_preset
      - fastq_minimap2_allow_secondary
      - igv_track_window_bp
      - igv_report_max_sites
      - igv_report_flanking_bp
      - enable_rotating_reference_frames
      - rotation_scan_step_bp
      - single_ref_split_min_mapq
      - single_ref_split_min_segment_bp
      - single_ref_split_max_query_gap_bp

  - id: pooled_reference_assignment
    name: ONT Pooled Reference Assignment
    description: Strict competitive assignment of one FASTQ against a server-staged immutable pooled reference set; ends at operator review.
    params:
      - fastq_path
      - reference_set_manifest
      - pooled_assignment_min_mapq
      - pooled_assignment_min_alignment_score_margin

  - id: methylation_analysis
    name: Methylation Analysis
    description: POD5/BAM methylation analysis with modkit
    params:
      - pod5_dir
      - bam_path
      - bam_force_realign
      - bam_min_mapq
      - reference_fasta
      - dorado_model
      - dorado_device
      - modified_bases
      - min_qscore
      - trim_adapters
      - emit_summary
      - dorado_batch_size
      - run_modkit
      - modkit_filter_threshold

  - id: clone_validation
    name: Clone Validation
    description: EPI2ME wf-clone-validation assembly with plasmid QC
    params:
      - pod5_dir
      - bam_path
      - fastq_path
      - reference_fasta
      - expected_plasmid_size
      - samtools_consensus_config
      - dorado_model
      - dorado_device
      - min_qscore
      - trim_adapters
      - emit_summary
      - wf_clone_sample
      - wf_clone_large_construct
      - wf_clone_approx_size
      - wf_clone_assm_coverage
      - wf_clone_trim_length
      - wf_clone_min_quality
      - wf_clone_assembly_tool
      - wf_clone_basecaller_model
      - wf_clone_primers
      - wf_clone_insert_reference
      - wf_clone_host_reference
      - wf_clone_regions_bedfile
      - wf_clone_flye_quality
      - wf_clone_non_uniform_coverage
      - wf_clone_canu_fast
      - wf_clone_cutsite_mismatch
      - wf_clone_primer_mismatch
      - wf_clone_expected_coverage
      - wf_clone_expected_identity
      - enable_rotating_reference_frames
      - rotation_scan_step_bp
      - single_ref_split_min_mapq
      - single_ref_split_min_segment_bp
      - single_ref_split_max_query_gap_bp

params:
  - name: pod5_dir
    type: directory
    description: Directory containing POD5 raw signal files
    required: false

  - name: bam_path
    type: file
    description: Existing BAM input (MM/ML supported)
    required: false
    file_type: bam

  - name: bam_force_realign
    type: boolean
    description: Realign existing BAM to reference via dorado aligner (disabled by default for BAM pass-through)
    required: false
    default: false

  - name: bam_min_mapq
    type: integer
    description: Minimum MAPQ retained when preparing/analyzing BAM input
    required: false
    default: 0
    minimum: 0
    maximum: 60

  - name: fastq_path
    type: file
    description: FASTQ(.gz) input for alignment-based plasmid QC
    required: false
    file_type: fastq

  - name: reference_fasta
    type: file
    description: Optional reference FASTA; required for modkit pileup BED output and optional BAM force-realignment
    required: false
    file_type: fasta

  - name: reference_set_manifest
    type: file
    description: Server-staged immutable bms.ngs.reference-set.v1 manifest for pooled assignment
    required: false
    file_type: json

  - name: pooled_assignment_min_mapq
    type: integer
    description: Minimum MAPQ retained as competitive pooled-assignment evidence
    required: false
    default: 20
    minimum: 0
    maximum: 255

  - name: pooled_assignment_min_alignment_score_margin
    type: integer
    description: Inclusive alignment-score margin within which competitors remain ambiguous
    required: false
    default: 10
    minimum: 0

  - name: dorado_model
    type: string
    description: Server-resolved exact Dorado model identifier
    required: false
    default: sup
    pattern: '^[A-Za-z0-9][A-Za-z0-9._@+-]{0,127}$'

  - name: dorado_quality_mode
    type: string
    description: Locked quality selector resolved server-side to an exact model identity
    required: false
    default: sup
    enum: [fast, hac, sup]

  - name: dorado_basecall_mode
    type: string
    description: Locked Dorado execution mode
    required: false
    default: simplex
    enum: [simplex, duplex]

  - name: ont_molecule_type
    type: string
    description: Molecule family used for chemistry/model qualification
    required: false
    default: dna
    enum: [dna, rna]

  - name: duplex_pairs
    type: file
    description: Confined explicit Dorado duplex read-pairs file
    required: false

  - name: barcode_kit
    type: string
    description: Locked inline barcode-classification kit
    required: false
    enum: [SQK-RBK114-96]

  - name: sample_sheet
    type: file
    description: Optional confined barcode sample sheet
    required: false

  - name: dorado_device
    type: string
    description: CUDA device selector passed to Dorado (e.g., cuda:0, cuda:all)
    required: false
    default: cuda:0
    pattern: '^(?:cuda:(?:all|auto|[0-9]+)|cpu|metal)$'

  - name: modified_bases
    type: string
    description: Explicit compatible modified-base model selection; none is the safe default
    required: false
    default: none
    enum:
      - 6mA
      - 5mC_5hmC
      - none

  - name: min_qscore
    type: integer
    description: Minimum read Q-score threshold for Dorado basecalling (POD5 mode)
    required: false
    default: 10
    minimum: 0
    maximum: 30

  - name: trim_adapters
    type: boolean
    description: Trim sequencing adapters during basecalling
    required: false
    default: true

  - name: emit_summary
    type: boolean
    description: Emit sequencing summary TSV from Dorado
    required: false
    default: true

  - name: dorado_batch_size
    type: integer
    description: Dorado inference batch size; 0 selects native auto, omitted uses simplex 64 or duplex 32
    required: false
    minimum: 0

  - name: run_modkit
    type: boolean
    description: Run modkit pileup and summary outputs
    required: false
    default: false

  - name: modkit_filter_threshold
    type: number
    description: Probability threshold for modkit filtering
    required: false
    default: 0.5
    minimum: 0.0
    maximum: 1.0

  - name: run_fastq_qc
    type: boolean
    description: Run FASTQ plasmid QC (alignment summary, coverage, consensus)
    required: false
    default: true

  - name: run_multimer_qc
    type: boolean
    description: Legacy alias for FASTQ QC enablement (kept for backward compatibility)
    required: false
    default: false

  - name: expected_plasmid_size
    type: integer
    description: "Expected plasmid length (bp): null selects Auto from the reference; a positive integer overrides it. Read-length ratios are length heuristics, not biological copy number."
    required: false
    default: null
    minimum: 1

  - name: min_fastq_read_length
    type: integer
    description: Ignore FASTQ reads shorter than this length before computing FASTQ QC metrics
    required: false
    default: 0
    minimum: 0

  - name: samtools_consensus_config
    type: string
    description: Samtools consensus preset; null preserves native Bayesian defaults. Choices are owned by the ONT fastq_qc parameter schema.
    required: false
    default: null

  - name: fastq_minimap2_preset
    type: string
    description: minimap2 preset used for FASTQ-to-reference alignment (bundled-runtime default is map-ont for minimap2 2.24 compatibility)
    required: false
    default: map-ont
    enum:
      - map-ont
      - map-hifi
      - map-pb
      - sr

  - name: fastq_minimap2_allow_secondary
    type: boolean
    description: Keep minimap2 secondary alignments in FASTQ alignment output
    required: false
    default: true

  - name: igv_track_window_bp
    type: integer
    description: Window size (bp) used to aggregate IGV analysis tracks for FASTQ QC
    required: false
    default: 100
    minimum: 1

  - name: igv_report_max_sites
    type: integer
    description: Maximum number of hotspot sites to include in generated IGV report artifacts
    required: false
    default: 40
    minimum: 1

  - name: igv_report_flanking_bp
    type: integer
    description: Flanking context (bp) around each IGV report hotspot site
    required: false
    default: 200
    minimum: 0

  - name: run_assembly
    type: boolean
    description: Run EPI2ME wf-clone-validation assembly and construct QC
    required: false
    default: false

  - name: wf_clone_sample
    type: string
    description: Optional sample name override for wf-clone-validation
    required: false

  - name: wf_clone_large_construct
    type: boolean
    description: Enable large construct mode in wf-clone-validation
    required: false
    default: false

  - name: wf_clone_approx_size
    type: integer
    description: Approximate plasmid size (bp) for wf-clone-validation
    required: false
    default: 7000
    minimum: 1

  - name: wf_clone_assm_coverage
    type: integer
    description: Assembly downsampling coverage for wf-clone-validation
    required: false
    default: 60
    minimum: 1

  - name: wf_clone_trim_length
    type: integer
    description: Number of bases to trim from both read ends before assembly
    required: false
    default: 0
    minimum: 0

  - name: wf_clone_min_quality
    type: integer
    description: Minimum read quality threshold for wf-clone-validation
    required: false
    default: 9
    minimum: 0
    maximum: 60

  - name: wf_clone_basecaller_model
    type: string
    description: Declared native Medaka consensus profile (not proof of read origin). Medaka 2.2.2 has no R10.4.1 FAST-trained resolver; select a scientifically appropriate explicit profile. Observed conflicts remain visible in input_model_provenance.json.
    required: false
    default: dna_r10.4.1_e8.2_400bps_hac@v6.0.0
    enum:
      - dna_r10.4.1_e8.2_400bps_hac@v6.0.0
      - dna_r10.4.1_e8.2_400bps_sup@v5.2.0
      - dna_r10.4.1_e8.2_400bps_hac@v5.2.0
      - dna_r10.4.1_e8.2_400bps_hac@v5.0.0
      - dna_r10.4.1_e8.2_400bps_sup@v5.0.0

  - name: wf_clone_primers
    type: file
    description: Optional primer TSV
    required: false

  - name: wf_clone_insert_reference
    type: file
    description: Insert FASTA; requires primers
    required: false

  - name: wf_clone_host_reference
    type: file
    description: Host FASTA for host-read filtering
    required: false

  - name: wf_clone_regions_bedfile
    type: file
    description: Host regions BED; requires host reference
    required: false

  - name: wf_clone_assembly_tool
    type: string
    description: Assembly backend to use for wf-clone-validation
    required: false
    default: flye
    enum:
      - flye
      - canu

  - name: wf_clone_flye_quality
    type: string
    description: Bounded Flye read-quality preset for clone validation
    required: false
    default: nano-hq
    enum: [nano-hq, nano-corr, nano-raw]

  - name: wf_clone_non_uniform_coverage
    type: boolean
    description: Enable the vendor non-uniform coverage mode when using Flye
    required: false
    default: false

  - name: wf_clone_canu_fast
    type: boolean
    description: Enable the vendor Canu fast mode when using Canu
    required: false
    default: false

  - name: wf_clone_cutsite_mismatch
    type: integer
    description: Allowed cut-site mismatches for clone validation
    required: false
    default: 1
    minimum: 0
    maximum: 10

  - name: wf_clone_primer_mismatch
    type: integer
    description: Allowed primer mismatches for clone validation
    required: false
    default: 2
    minimum: 0
    maximum: 10

  - name: wf_clone_expected_coverage
    type: number
    description: Expected reference coverage percentage for clone validation
    required: false
    default: 95
    minimum: 0
    maximum: 100

  - name: wf_clone_expected_identity
    type: number
    description: Expected reference identity percentage for clone validation
    required: false
    default: 99
    minimum: 0
    maximum: 100

  - name: enable_rotating_reference_frames
    type: boolean
    description: Evaluate circular-reference rotations during plasmid dimer QC
    required: false
    default: true

  - name: rotation_scan_step_bp
    type: integer
    description: Rotation scan interval in base pairs for plasmid dimer QC
    required: false
    default: 1
    minimum: 1
    maximum: 10000

  - name: single_ref_split_min_mapq
    type: integer
    description: Minimum MAPQ for single-reference split evidence
    required: false
    default: 20
    minimum: 0
    maximum: 60

  - name: single_ref_split_min_segment_bp
    type: integer
    description: Minimum segment length for single-reference split evidence
    required: false
    default: 250
    minimum: 1
    maximum: 1000000

  - name: single_ref_split_max_query_gap_bp
    type: integer
    description: Maximum query gap for single-reference split evidence
    required: false
    default: 500
    minimum: 0
    maximum: 1000000

enabled: true
experimental: true