Skip to content

Cancer Genomics

Tumor genomes from the WiGiTS toolkit: sequencing QC, purity and ploidy, copy number, structural variants, the driver catalog, somatic and protein-level variants, and immune context.

nf-co.re GitHub

Experimental
Template version latest

The oncoanalyser template reads the tables the Hartwig Medical Foundation WiGiTS toolkit writes for each tumor, from sequencing QC to the protein changes and neoepitopes of its variants:

  • Overview: the run in four key figures, live findings and four figures, each linked to the tab that explains it
  • Data & QC: BamTools coverage and insert sizes, and QSEE's verdict on each check
  • Tumor profile: purity and ploidy, CUPPA's tissue of origin, COSMIC signatures and the HRD probability
  • Copy number and structural variants: PURPLE's copy-number profile, gene-level events, and LINX's rearrangements and fusions
  • Drivers and variants: the driver catalog, somatic small variants, and protein changes on the sequence and on a predicted structure
  • Immune context: HLA alleles, candidate neoepitopes and telomere length
Live showcase nf-core/oncoanalyser, nf-core test data Open

Tables only

The template reads the tools' tables and the PURPLE somatic VCF. It reads no BAM and no MultiQC report (the pipeline writes none), so it has no MultiQC tab. Isofox, PEACH, CIDER, ORANGE and the virus tables are not read yet.

Quick start

depictio ingest /path/to/oncoanalyser_results \
  --template nf-core/oncoanalyser/latest

The pipeline outdir is what you pass: one <group_id>/ directory per tumor, plus pipeline_info/. oncoanalyser does not publish the samplesheet it ran on, so the template reads input/samplesheet.csv under that directory by default. Copy your --input sheet there, or name it. A sheet you name makes the CLI guess the id and group columns from its header, so pin them:

depictio ingest /path/to/oncoanalyser_results \
  --template nf-core/oncoanalyser/latest \
  --var METADATA_FILE=/path/to/samplesheet.csv \
  --var METADATA_ID_COL=sample_id \
  --var GROUP_COL=sample_class
Variable Default Role
METADATA_FILE {DATA_ROOT}/input/samplesheet.csv The samplesheet the run started with
GROUP_COL sample_class Sheet column the dashboard groups samples by. The default is sample_type and sequence_type joined, e.g. tumor_dna
GENOME hg38 UCSC name of the assembly: hg38 for GRCh38_hmf, hg19 for GRCh37_hmf. It sets the chord diagram, the genome track and its gene lane 
depictio config nextflow --install     # once per machine
nextflow run nf-core/oncoanalyser -r 3.0.0 -profile docker \
  --mode wgts --genome GRCh38_hmf --input samplesheet.csv --outdir results

No depictio ingest, and no template named: the pipeline ingests its own output directory when it finishes and resolves this template from its own manifest. See Nextflow trigger. The trigger passes no --var, so put a copy of the samplesheet at results/input/samplesheet.csv before the run, and ingest by hand for a --genome GRCh37_hmf run.

Reference

The template reads the per-tumor tables of PURPLE, LINX, BamTools, QSEE, CUPPA, SIGS, CHORD, LILAC, TEAL and Neo, and PURPLE's somatic VCF, of which it keeps the PASS calls. Every output sits under <group_id>/<tool>/, so a cohort run is a set of group directories side by side, and each recipe globs for its files. The protein changes come from the PAVE annotation of the VCF: the three-letter residues of each HGVS protein change become a residue number and one-letter amino acids. A driver row from LINX is kept only when PURPLE has no row for the same tumor, gene and driver type, so LINX's disruptions are added without doubling the reported drivers.

Optional tools drop their tiles

CUPPA, SIGS, CHORD, LILAC, TEAL, Neo and QSEE can be excluded with --processes_exclude, and a targeted (panel) run skips several of them. A run without them skips those collections at ingest, and the dashboard drops the tiles that bind them: the Findings row and the Pipeline step that cite them, a section whose tiles all go, and the Immune context tab when LILAC, Neo and TEAL are all missing. Without QSEE only the QC verdicts and the QSEE table go.

Pipeline nf-core/oncoanalyser 3.0.0 Template 2.1.0 Updated 1 Nov 2026

Template variables

Variables you provide when running the template: DATA_ROOT is the results directory given to depictio ingest, the others are set with --var NAME=value:

Variable Required Description
DATA_ROOT ✓ oncoanalyser output directory (one / sub-directory per tumor, plus pipeline_info/).
METADATA_FILE — The samplesheet the run was started with (--input: group_id, subject_id, sample_id, sample_type, sequence_type, filetype, filepath). Defaults to input/samplesheet.csv under DATA_ROOT.
METADATA_ID_COL — Samplesheet id column. oncoanalyser fixes it to sample_id, the name every per-sample output file carries.
GROUP_COL — Samplesheet column the dashboard groups and filters samples by. Defaults to sample_class (sample_type and sequence_type joined, e.g. tumor_dna), which the sample hub derives.
GROUP_COL_DISPLAY — Reader-facing label of GROUP_COL, used in filter and card titles.
GENOME — Genome build of the run (UCSC name): hg38 for GRCh38_hmf, hg19 for GRCh37_hmf. Sets the chromosome ring, the locus views and their gene lane.
TRACKS_URI — s3:// or https:// folder holding the run's outdir (the one with a / sub-directory per tumor), to read the genome-browser tracks (PURPLE's VCFs and the BAMs) in place instead of uploading them from DATA_ROOT. Without it the tracks collection uploads whichever of those files DATA_ROOT holds.

Data collections

23 data collections — 15 required 8 optional · 20 direct 3 derived.

Origin tells you whether a collection is real pipeline data or a reshape of it: direct = a pipeline output (scanned, or a recipe that reads raw files); derived = a recipe that reshapes one or more direct collections into the layout a visualization needs (no new measurement). Reads shows what produces it: a recipe .py transform, whose name links to its source on GitHub, or a raw file scanned off disk. (A direct collection can still have a recipe — one that merely parses/cleans the raw file; derived means the recipe reshapes another collection.)

Tag Origin Type Reads Status
samples direct Table recipe nf-core/oncoanalyser/samples.py required
tracks direct genomic_tracks recipe nf-core/oncoanalyser/tracks.py required
purple_purity direct Table recipe purple/purity.py required
purple_cnv_segments direct Table recipe purple/cnv_segments.py required
purple_gene_copy_number direct Table recipe purple/gene_copy_number.py required
purple_drivers direct Table recipe purple/drivers.py required
purple_somatic_variants direct Table recipe purple/somatic_variants.py required
purple_protein_changes derived Table recipe purple/protein_changes.py required
purple_protein_genes derived Table recipe nf-core/oncoanalyser/protein_genes.py required
linx_svs direct Table recipe linx/svs.py required
linx_fusions direct Table recipe linx/fusions.py required
linx_fusion_structure derived Table recipe linx/fusion_structure.py required
cuppa_predictions direct Table recipe cuppa/predictions.py optional
cuppa_classifier_matrix direct Table recipe cuppa/classifier_matrix.py optional
sigs_allocation direct Table recipe sigs/allocation.py optional
chord_prediction direct Table file .*.chord.prediction.tsv$ optional
teal_telomere_length direct Table file .*.teal.tellength.tsv$ optional
lilac_alleles direct Table recipe lilac/alleles.py optional
neo_neoepitopes direct Table recipe neo/neoepitopes.py optional
hmf_bamtools_summary direct Table recipe hmf_bamtools/summary.py required
hmf_bamtools_coverage_profile direct Table recipe hmf_bamtools/coverage_profile.py required
hmf_bamtools_fragment_length direct Table recipe hmf_bamtools/fragment_length.py required
qsee_status direct Table recipe qsee/status.py optional

Conditional routes

Rows are data collections; columns are the variables you set or params.json flags auto-detected from the run. Each filled cell is the effect of setting that variable; an empty cell means that variable leaves the collection unchanged. (22 collections are unaffected by any variable — present on every run.)

+ included− removed⇄ repointed

Data collection TRACKS_URI
tracks ⇄

28 links — selecting a value in the source collection filters the target. The join column is shown after the source.

Source · column Target Filters
samples · sample_id → purple_purity Filter the purity fit by samplesheet samples
samples · sample_id → purple_cnv_segments Filter the copy-number segments by samplesheet samples
samples · sample_id → purple_gene_copy_number Filter the gene copy number by samplesheet samples
samples · sample_id → purple_drivers Filter the driver catalog by samplesheet samples
samples · sample_id → purple_somatic_variants Filter the somatic variants by samplesheet samples
samples · sample_id → purple_protein_changes Filter the protein changes by samplesheet samples
samples · sample_id → purple_protein_genes Filter the genes with a protein change by samplesheet samples
samples · sample_id → linx_svs Filter the structural variants by samplesheet samples
samples · sample_id → linx_fusions Filter the fusions by samplesheet samples
samples · sample_id → linx_fusion_structure Filter the fusion structures by samplesheet samples
samples · sample_id → cuppa_predictions Filter the tissue-of-origin predictions by samplesheet samples
samples · sample_id → cuppa_classifier_matrix Filter the classifier matrix by samplesheet samples
samples · sample_id → sigs_allocation Filter the signature allocation by samplesheet samples
samples · sample_id → chord_prediction Filter the HRD prediction by samplesheet samples
samples · sample_id → teal_telomere_length Filter the telomere lengths by samplesheet samples
samples · sample_id → lilac_alleles Filter the HLA alleles by samplesheet samples
samples · sample_id → neo_neoepitopes Filter the neoepitopes by samplesheet samples
samples · sample_id → hmf_bamtools_summary Filter the alignment summary by samplesheet samples
samples · sample_id → hmf_bamtools_coverage_profile Filter the coverage curves by samplesheet samples
samples · sample_id → hmf_bamtools_fragment_length Filter the fragment length curves by samplesheet samples
samples · sample_id → qsee_status Filter the QC verdicts by samplesheet samples
purple_drivers · gene → purple_protein_changes Filter the protein changes by the genes of the selected drivers
purple_gene_copy_number · gene → purple_drivers Filter the driver catalog by the selected genes
samples · sample_id → tracks Show the genome tracks of the samples picked in the left panel
tracks · sample_id → samples Filter the dashboard by the samples whose tracks are picked in the genome browser
purple_somatic_variants · sample → tracks Open the genome tracks of the tumor a picked call was made in
purple_cnv_segments · sample → tracks Open the genome tracks of the tumor a picked copy-number segment belongs to
linx_svs · sample → tracks Open the genome tracks of the tumor a picked structural variant was called in

Recipes

Each recipe reshapes raw pipeline output into a tidy table. The name links to its source; Output lists the validated OUTPUT_SCHEMA columns.

Recipe Transforms Output
cuppa/classifier_matrix.py CUPPA's probabilities as a cancer type by classifier matrix, for a heatmap. row_label, sample, cancer_type, combined
cuppa/predictions.py CUPPA's tissue-of-origin probabilities, one row per tumor, classifier and cancer type. sample, classifier_group, classifier, cancer_type, probability, rank, is_top
hmf_bamtools/coverage_profile.py hmftools BamTools coverage distribution, as a cumulative curve per sample. sample, coverage, bases, fraction, fraction_at_least
hmf_bamtools/fragment_length.py hmftools BamTools fragment length distribution, decimated for a profile. sample, fragment_length, pairs, fraction
hmf_bamtools/summary.py hmftools BamTools alignment summary, one row per DNA sample. sample, region_bases, total_reads, duplicate_reads, dual_strand_reads, mean_coverage, sd_coverage, median_coverage, mad_coverage, low_map_qual_frac, duplicate_frac, unmapped_frac, low_base_qual_frac, overlapping_read_frac, capped_coverage_frac, frac_ge_10x, frac_ge_20x, frac_ge_30x, frac_ge_60x
lilac/alleles.py LILAC's HLA class I and II typing, one row per tumor and called allele. allele_id, sample, gene, allele, ref_fragments, ref_unique_fragments, tumor_fragments, rna_fragments, tumor_copy_number, lost_in_tumor, somatic_missense, somatic_nonsense_or_frameshift, somatic_splice, somatic_inframe_indel, somatic_synonymous, somatic_mutations
linx/fusion_structure.py LINX fusions as two partner lanes along the fused transcript, in exon units. fusion_id, fusion, sample, partner, feature, start, end, retained, breakpoint, phased, reported
linx/fusions.py LINX's predicted gene fusions, one row per fusion, with both partners resolved. fusion_id, sample, fusion, gene_up, gene_down, reported, reported_type, reportable_reasons, phased, likelihood, transcript_up, transcript_down, region_up, region_down, coords_up, coords_down, fused_exon_up, fused_exon_down, exons_up_total, exons_down_total, skipped_exons_up, skipped_exons_down, chain_length, chain_links, chain_terminated, domains_kept, domains_lost
linx/svs.py LINX's annotated somatic structural variants, one row per SV, as locus pairs. sv_key, sample, sv_id, label, type, cluster_id, resolved_type, cluster_category, cluster_sv_count, chrom_a, pos_a, orient_a, chrom_b, pos_b, orient_b, length, junction_copy_number, junction_copy_number_min, junction_copy_number_max, genes_a, genes_b, is_foldback, fragile_site, line_element
neo/neoepitopes.py Neo's scored neoepitopes, one row per tumor and candidate neoepitope. neoepitope_id, sample, variant_type, variant, gene, up_amino_acids, novel_amino_acids, down_amino_acids, peptides, rna_fragments, rna_depth, expected_tpm, effective_tpm, variant_copy_number, copy_number, subclonal_likelihood
nf-core/oncoanalyser/protein_genes.py One row per tumor and gene with a protein change: the protein tab's selector. entity, sample, changes, hotspots, max_af, worst_effect, driver_likelihood, gene_role, driver_types, reported
nf-core/oncoanalyser/samples.py Sample hub for nf-core/oncoanalyser: the run's samplesheet, one row per sample. sample_id, group_id, subject_id, sample_type, sequence_type, sample_class, input_files, filetypes
nf-core/oncoanalyser/tracks.py Genome-browser track manifest: one row per track file the run published. track_id, sample_id, group_id, source_sample, kind, format, uri, name, category, order
purple/cnv_segments.py PURPLE's fitted copy-number segments as a canonical cnv_profile table. sample, chrom, start, end, log2, baf, copy_number, segment, label, depth, minor_copy_number, major_copy_number
purple/drivers.py The run's driver catalog, somatic and germline, one row per gene and driver. driver_id, sample, gene, driver, origin, source_tool, category, reported, report_status, driver_likelihood, likelihood_method, biallelic, missense, nonsense, splice, inframe, frameshift, min_copy_number, max_copy_number, chrom, chromosome_band, transcript
purple/gene_copy_number.py PURPLE's copy number per gene, with a call a reader can filter on. sample, gene, chrom, start, end, chromosome_band, transcript, min_copy_number, max_copy_number, min_minor_allele_copy_number, relative_min_copy_number, somatic_regions, cn_event, reported_status, driver_type
purple/protein_changes.py Protein-changing somatic variants in the protein module's variant contract. entity, position, ref_aa, alt_aa, category, value, label, sample, hgvs_p, hgvs_c, transcript, canonical_effect, variant_key, chrom, pos, variant_copy_number, biallelic, hotspot, reported
purple/purity.py PURPLE's tumor fit and QC verdict, one row per tumor sample. sample, purity, ploidy, fit_status, sex, whole_genome_duplication, ms_indels_per_mb, ms_status, tmb_per_mb, tmb_status, tml, tml_status, sv_tmb, diploid_fraction, polyclonal_fraction, min_purity, max_purity, min_ploidy, max_ploidy, fit_score, run_mode, qc_status, fit_method, cn_segments, unsupported_cn_segments, deleted_genes, contamination, germline_aberrations, amber_mean_depth, loh_fraction, tinc_level
purple/somatic_variants.py PURPLE's somatic small variants, one row per PASS call, with PAVE's impact. variant_key, sample, chrom, pos, ref, alt, variant_type, tier, gene, transcript, canonical_effect, coding_effect, hgvs_c, hgvs_p, worst_coding_effect, splice_region, tumor_af, tumor_depth, tumor_alt_reads, purple_af, copy_number, variant_copy_number, minor_allele_copy_number, subclonal_likelihood, biallelic, hotspot, reported, kataegis, trinucleotide_context, germline_status, gnomad_frequency, mappability
qsee/status.py QSEE's QC verdicts, one row per sample and checked feature. check_id, sample, sample_type, source_tool, feature_type, feature, value, qc_status, fail_condition
sigs/allocation.py SIGS' COSMIC signature allocation, one row per tumor and signature. sample, signature, allocation, fraction, percent

Dashboard tabs

One dashboard: the Overview, then eight child tabs in three groups, read as a funnel from the sequencing to the tumor's genome and then to its variants. Each tab below carries the same icon and colour the dashboard gives it, so the page and the app read alike.

Group Tabs
Data & QC Sequencing QC
Genome Tumor profile, Copy number, Structural variants
Variants Drivers, Small variants, Protein changes, Immune context

Each child tab opens with a short intro and a strip of four cards, then at most three open sections; tables and record cards follow, collapsed. The persistent Sample filters (sample class, sample, analysis group) sit in the left panel and narrow every tab through the samplesheet links. The Sample sheet is pinned, collapsed, to the bottom of every child tab. The Structural variants and Small variants tabs each embed a genome browser with the tumor's calls and reads.

Tumor genomes, from the reads to purity, copy number, drivers, variants and immune context.

Overview dashboard

A short hero links the run parameters. About this dashboard says how the two filter levels work, The run lists the samples, analysis groups, mode, genome and purity fit, and Pipeline walks six steps (align, fit, link, call, classify, type), each linked to its settings and its tab. The findings are live values and follow the filters: the leading signature's share of SNVs, genes lost and gained, structural variants by their commonest resolved event, and somatic calls per megabase. A run without CUPPA and SIGS loses the signature finding.

Filters and components

Filters · the left panel starts collapsed. Key figures and Findings each have their own filter bar (sample class and sample id): each narrows its own section only.

Section What it holds
Top Hero, About this dashboard, The run, Pipeline
Key figures 4 headline cards: samples, tumor purity, reported drivers, mean coverage
Findings Live result rows, then 4 figures: the signature shares, the copy-number profile, the structural variant chord diagram and the allele frequency histogram
How to read this dashboard The tabs by group, each with its question

Data & QC · Is the sequencing deep and clean enough for the calls?

BamTools' summary and QSEE's checks. The mean coverage, reads per sample, the highest duplicate share and the lowest share at 30x or more. Then the cumulative coverage and the insert size distribution, one curve per sample, and QSEE's PASS, WARN and FAIL verdicts per tool. The BamTools summary with a library record and the QSEE check table are collapsed.

Filters and components

Filters · a Mean coverage range on hmf_bamtools_summary, and QC verdict and Source tool on qsee_status.

Section What it holds
QC at a glance 4 cards
Coverage and insert size 2 advanced visualizations
QC verdicts QC verdicts per tool
Library detail (collapsed) BamTools summary per sample, Library record
QC reference tables (collapsed) QSEE checks

Genome · How pure is the tumor, and what kind of tumor is it?

PURPLE's fit and the classifiers around it. Purity, ploidy, mutations per Mb and microsatellite indels. Then CUPPA's cancer-type probability per classifier as a heatmap, the share of SNVs per COSMIC signature beside CHORD's HRD probability and the SNVs allocated, and purity against ploidy with one point per tumor. The PURPLE fit table with a tumor record and the CUPPA and CHORD tables are collapsed.

Filters and components

Filters · PURPLE fit status and PURPLE QC on purple_purity.

Section What it holds
Tumor at a glance 4 cards
Tissue of origin 1 advanced visualization: the CUPPA heatmap
Mutational processes Share of SNVs per mutational signature, 2 cards
Purity and ploidy 1 advanced visualization
Tumor detail (collapsed) PURPLE fit per tumor, Tumor record
Reference tables (collapsed) CUPPA probabilities, CHORD prediction

Genome · Which regions are gained or lost, and in which genes?

Segments, the median segment copy number, genes in view and the LOH share of the genome. Then PURPLE's copy-number profile along the genome (log2 of copy number over 2, with the B-allele frequency underneath) and the lowest against the highest copy number per gene. Events follow PURPLE's cut-offs: AMP when the lowest copy number is above three times the ploidy, DEL below 0.5 copies, LOH when the minor allele is below 0.5 copies, and PARTIAL_* when only part of the gene qualifies. The tab opens with the event filter on whole and partial amplifications and deletions so the gene scatter stays readable; clearing it adds the genes with LOH or no change. The gene table with a gene record is collapsed.

Filters and components

Filters · Gene copy-number event on purple_gene_copy_number and Chromosome on purple_cnv_segments.

Section What it holds
Copy number at a glance 4 cards
Copy-number profile 1 advanced visualization
Gene copy number 1 advanced visualization
Gene detail (collapsed) Copy number per gene, Gene record

Genome · Which rearrangements shaped the genome, and which fusions follow?

LINX's clustered rearrangements and fusion calls. Structural variants, SV clusters, the median junction copy number and candidate fusions. Then a chord diagram of the breakend pairs beside the SVs per resolved event, and the kept and lost exons per fusion beside the fusions per reading frame. The pipeline writes no transcript model, so exon spans are drawn in exon units along the fused transcript, from the fused and total exon counts of each partner. The fusion table with a fusion record and the SV table are collapsed.

Filters and components

Filters · Resolved event and SV type on linx_svs, and Fusion phase on linx_fusions.

Section What it holds
Structural variants at a glance 4 cards
Rearrangement landscape 1 advanced visualization: the chord diagram, and SVs per resolved event
Fusion structure 1 advanced visualization: the exon layout, and Fusions per reading frame
Fusion detail (collapsed) Candidate fusions, Fusion record
Structural variant tables (collapsed) Structural variants

Variants · Which genes drive this tumor, and how?

PURPLE's driver catalog, with LINX's disruptions added. Driver entries, oncogenes and suppressors, biallelic entries and the median driver likelihood. Then an oncoplot of driver genes by tumor and driver type beside the entries per gene role, and the likelihood per driver type. The tab opens with Report status on REPORTED so the oncoplot stays readable on a whole-genome catalog; clearing it adds the unreported entries. The driver table with a driver record is collapsed.

Filters and components

Filters · Driver type, Gene role, Origin and Report status on purple_drivers.

Section What it holds
Drivers at a glance 4 cards
Driver landscape 1 advanced visualization: the oncoplot, and Entries per gene role and driver type
Driver likelihood Driver likelihood per driver type
Driver detail (collapsed) Driver catalog, Driver record

Variants · How many somatic variants are there, and how clonal?

PASS calls of PURPLE's somatic VCF, with the gene, effect and HGVS change from PAVE's canonical transcript. Somatic calls, coding calls, the median purity-adjusted allele frequency and the median variant copy number. Then their histograms (a variant type keeps one colour on every tile) and every call along the genome, with a gene lane. The variant table with a variant record is collapsed.

Filters and components

Filters · Coding effect, Variant type and a Purity-adjusted allele frequency range on purple_somatic_variants.

Section What it holds
Variants at a glance 4 cards
Clonality Purity-adjusted allele frequency, Variant copy number
Along the genome 1 advanced visualization
Variant detail (collapsed) Somatic small variants, Variant record

Variants · Which proteins change, and where on the protein and its structure?

Protein changes dashboard

Protein changes, proteins hit, changes at known hotspots and the median allele frequency. Then a scatter of the genes with a change (highest purity-adjusted allele frequency against PURPLE's driver likelihood, sized by change count, coloured by worst effect) beside the predicted structure of the picked gene over its written sequence, and a full-width lollipop of its changes. Click a gene on the scatter and the other two follow it; click a residue in one and it is marked in the others, red on the structure. Without a pick the views open on the protein with the most changes, and the Protein filter drives them too. The change table with a record is collapsed.

The structure tile needs the resolver

The pipeline predicts no structure. The 3D structure tile looks the picked gene up by its symbol (UniProt, then the AlphaFold DB model), which needs DEPICTIO_STRUCTURE_RESOLVER_ENABLED=true on the server (it is off by default; see the structure resolver). With it off the tile says the lookup is disabled, and the scatter, the lollipop and the table keep working. A change whose reference residue disagrees with the model is listed, not drawn.

Filters and components

Filters · Protein (a single pick) and Coding effect on purple_protein_changes.

Section What it holds
Protein changes at a glance 4 cards
Protein view 3 advanced visualizations: the gene scatter, the structure and the lollipop
Protein change detail (collapsed) Protein changes, Protein change record

Variants · Which HLA alleles does the tumor keep, and what could they present?

LILAC's HLA alleles, Neo's candidate neoepitopes and TEAL's telomere length. HLA alleles, candidate neoepitopes, their median expression and telomere length. Then the tumor copy number per HLA allele beside the telomere length per sample, and neoepitope expression against variant copy number beside the neoepitopes per source event. The neoepitope table with a record and the HLA allele and telomere tables are collapsed. The tab is dropped when LILAC, Neo and TEAL are all missing.

Filters and components

Filters · HLA gene on lilac_alleles and Neoepitope source on neo_neoepitopes.

Section What it holds
Immune context at a glance 4 cards
HLA and telomeres Tumor copy number per HLA allele, Telomere length per sample
Neoepitopes 1 advanced visualization, Neoepitopes per source event
Neoepitope detail (collapsed) Candidate neoepitopes, Neoepitope record
Immune reference tables (collapsed) HLA alleles, Telomere length

The sample sheet selects on sample_id, and the sample links carry the pick to every collection. A driver pick narrows the protein changes to its genes, and a gene picked in the copy-number table or scatter narrows the driver catalog. Each record card sits beside the table that drives it and opens on the selected row.

Running the pipeline

Depictio reads the output of nf-core/oncoanalyser, it does not run the pipeline. Run the pipeline first; its --input sheet has one row per input file with group_id, subject_id, sample_id, sample_type and sequence_type:

nextflow run nf-core/oncoanalyser -r 3.0.0 \
  --mode wgts \
  --genome GRCh38_hmf \
  --input samplesheet.csv \
  --outdir results -profile docker

Then point Depictio at the results, with the samplesheet the run started with:

depictio ingest results/ \
  --template nf-core/oncoanalyser/latest \
  --var METADATA_FILE=samplesheet.csv \
  --var METADATA_ID_COL=sample_id \
  --var GROUP_COL=sample_class

See nf-co.re/oncoanalyser/usage for full pipeline documentation.

Required data structure

Point depictio ingest at the pipeline outdir. Depictio scans recursively and matches on file name; every tumor-level file is named after the tumor's sample_id, which is the key the sample hub links on.

<DATA_ROOT>/
├── input/samplesheet.csv                          # not published by the pipeline: copy it here, or --var METADATA_FILE
├── pipeline_info/
│   ├── params*.json
│   └── *software*versions*.yml
└── <group_id>/                                    # one directory per tumor
    ├── purple/
    │   ├── *.purple.{purity.tsv,qc,cnv.somatic.tsv,cnv.gene.tsv,somatic.vcf.gz}
    │   └── *.driver.catalog.*.tsv
    ├── linx/
    │   ├── somatic_annotations/*.linx.{svs,clusters,fusion,breakend}.tsv
    │   └── {somatic,germline}_annotations/*.driver.catalog.tsv
    ├── bamtools/<sample>/*.bam_metric.{summary,coverage,frag_length}.tsv
    ├── qsee/*.qsee.status.tsv.gz                  # optional
    ├── cuppa/*.cuppa.vis_data.tsv                 # optional
    ├── sigs/*.sig.allocation.tsv                  # optional
    ├── chord/*.chord.prediction.tsv               # optional
    ├── lilac/*.lilac.tsv                          # optional
    ├── teal/*.teal.tellength.tsv                  # optional
    └── neo/scorer/*.neo.neoepitope.tsv            # optional

Validation runs

The template was validated against the AWS megatest of the 3.0.0 release, results-7c74c87a43749952b38c9a18915947570f0595a0, on its test_full profile: one WGTS subject, a tumor and normal DNA pair plus tumor RNA. The full run is 272 GB, almost all of it BAMs. megatest.yaml lists the tables-only subset the template needs, about 33 MB, and the script copies the samplesheet under input/. The screenshots above come from that run:

bash depictio/projects/nf-core/oncoanalyser/3.0.0/download_test_data.sh /tmp/oncoanalyser_test
depictio ingest /tmp/oncoanalyser_test --template nf-core/oncoanalyser/latest

Do not pass --project when ingesting: the dashboard is attached to the project by name, so renaming it breaks a later depictio dashboard import. Re-ingesting accumulates dashboards, so delete the project before repeating a run.

Additional resources

Authorship

Developers Wrote the template, its recipes and its dashboards. weber8thomas
Reviewers Nobody has run it on their own data and signed it off yet, which is what keeps it from Reviewed. Open
Maintainers Keep it working as nf-core/oncoanalyser releases. weber8thomas

Reviewing a template on your own data, or taking over a role here, is a contribution in itself: the contributing guide says what each one involves.