Foundry
Home Pattern

Interval: compute coverage

Two coverage modes: genome-wide depth as a bedgraph (genomecoveragebed) and reads counted in given regions (coveragebed). Same family, different question.

draft patterntarget/galaxytopic/galaxy-transformtopic/interval-transform
Revised
2026-06-10
Rev
1
Patterns
Molds
Related

Pattern health

warn
  • IWC exemplar anchors

    3 abstract workflow anchors declared.

  • Foundry verification fixture

    No structural verification fixture yet.

  • Pattern map coverage

    1 pattern map link here.

  • Metadata contract

    Pattern frontmatter matches the site contract.

Interval: compute coverage

Operation

“Coverage” in the IWC corpus is two operations sharing the bedtools family and the word. Decide which question you are asking before picking the tool.

  • Genome-wide depth → bedgraph. toolshed.g2.bx.psu.edu/repos/iuc/bedtools/bedtools_genomecoveragebed (“bedtools Genome Coverage”) emits per-position or per-interval depth across the whole genome. Used as a normalized coverage track (transcriptomics) or as the input to a low-coverage mask (SARS-CoV-2 consensus).
  • Reads in given regions → counts. toolshed.g2.bx.psu.edu/repos/iuc/bedtools/bedtools_coveragebed (“bedtools Coverage”) counts/measures how much of a second dataset (reads, features) falls inside each region of a supplied interval set.

Parameter names are corpus-inferred from tool_state.

When to reach for it

  • Track or mask input (genomecoveragebed): you have alignments (BAM) and want depth everywhere — to visualize, or to threshold into low/high-coverage regions. Feeds interval-mask-by-set-algebra.
  • Reads-in-regions (coveragebed): you already have the regions of interest (peaks, windows, a panel) and want a number per region. This is the quantify step of interval-windowed-coverage.

If you only need a yes/no overlap, not a count, use interval-overlap-filter.

Mode A — genome-wide depth (genomecoveragebed)

  • input_type.input_type_select: bam (corpus) or bed.
  • report.report_select: bg (bedgraph) in every corpus case.
    • report.zero_regions: true emits zero-depth spans as explicit rows — required when a downstream filter must see the gaps (the mask use, consensus-from-variation). false omits them (the track use, rnaseq-pe).
    • report.scale: a multiplier; a connected ConnectedValue carrying 1/million-reads gives a normalized track (rnaseq-pe). "1.0" for raw depth.
  • split: true splits spliced/blocked alignments into their blocks before counting (corpus: true for the mask case).
  • strand: "" (both strands) in the corpus.
  • d, dz, five, three: per-position / 5’/3’ report flags (bedtools -d/-dz/-5/-3); all false in the corpus (bedgraph mode is used instead).
tool_id: toolshed.g2.bx.psu.edu/repos/iuc/bedtools/bedtools_genomecoveragebed/2.31.1
tool_state:
  input_type: { input_type_select: bam, input: { __class__: ConnectedValue } }
  report: { report_select: bg, zero_regions: true, scale: "1.0" }
  split: true
  strand: ""
  d: false
  dz: false
  five: false
  three: false

Mode B — reads in given regions (coveragebed)

  • inputA: the regions to quantify over (connected).
  • reduce_or_iterate: iterate in the corpus — inputB (the reads/features dataset) applied per collection element.
  • hist: true reports a coverage histogram per region; false (corpus) reports per-region summary columns.
  • mean: true reports mean depth instead of counts; false in the corpus.
  • d, a_or_b, overlap_a, overlap_b, reciprocal_overlap, sorted, split, strandedness: report/overlap toggles; corpus defaults (false/null).
tool_id: toolshed.g2.bx.psu.edu/repos/iuc/bedtools/bedtools_coveragebed/2.31.1+galaxy0
tool_state:
  inputA: { __class__: ConnectedValue }
  reduce_or_iterate:
    reduce_or_iterate_selector: iterate
    inputB: { __class__: ConnectedValue }
  hist: false
  mean: false
  d: false
  sorted: false
  split: false
  strandedness: false

Pitfalls

  • Picking the wrong mode. genomecoveragebed answers “how deep is every position?”; coveragebed answers “how much falls in these regions?” They are not interchangeable — coveragebed needs you to supply the regions; genomecoveragebed invents them from the genome.
  • zero_regions is correctness, not cosmetics, for masking. Omit it and zero-depth spans never appear as rows, so a c4 < threshold filter cannot select them — the mask silently misses fully-uncovered regions. See interval-mask-by-set-algebra.
  • split matters for spliced reads. RNA-seq/coverage over spliced alignments without split: true counts the intron span as covered. The corpus sets split: true wherever spliced input is plausible.
  • bedgraph is interval data, but downstream often treats it as tabular. A Filter1 c4 < N on the bedgraph is a tabular filter on coordinate data — the interval↔tabular seam; see galaxy-tabular-patterns.

See also

IWC exemplars3 anchors

IWC Exemplars

sars-cov-2-variant-calling/sars-cov-2-consensus-from-variation/consensus-from-variationhigh

genomecoveragebed as a bedgraph with zero_regions, feeding a depth-threshold mask filter.

  • Depth of coverage
transcriptomics/rnaseq-pe/rnaseq-pehigh

genomecoveragebed as a normalized (1/million-reads scale) bedgraph coverage track.

  • Scaled Coverage both strands combined
epigenetics/atacseq/atacseqhigh

coveragebed counting reads from a BAM that fall inside merged summit windows, in iterate mode.

  • Compute coverage on summits +/-500kb

Incoming References (6)

  • Galaxy: genomic interval patternsrelated pattern— Use this MOC to choose corpus-grounded Galaxy genomic interval operations and recipes on coordinate features.
  • Interval: build a mask by set algebrarelated pattern— Compute regions from regions: concatenate candidate intervals, merge into non-overlapping spans, then subtract the set to keep. The gops_* set-algebra recipe.
  • Interval: filter or annotate by overlaprelated pattern— Keep, drop, or annotate coordinate features by overlap with a second feature set; bedtools intersect (BED) or vcfvcfintersect (VCF), mapped over a collection.
  • Interval: window or flank featuresrelated pattern— Extend features by a fixed or fractional amount to build neighborhood windows, clamped to chromosome ends; bedtools slopbed with a genome file.
  • Interval: windowed coverage around featuresrelated pattern— Quantify signal in fixed neighborhoods around point features: window the features (slop), collapse overlaps (merge), then count reads in each window (coverage).
  • Iwc Interval Operations Surveyrelated note— IWC corpus survey of coordinate-aware genomic interval operations; sizing and candidate boundaries for a galaxy-interval-patterns MOC, with hold-if-thin gate.