Use caseDNA and genomesResearch and clinical research

Select a tumour DNA somatic variant-calling workflow

Which calling or rescoring workflow reliably detects tumour SNVs and small indels under the available sequencing and normal-sample regime?

In this release 1 evaluated endpoint · 2 tested configurations · Direct evidence for the stated endpoint · 3 recorded evidence gaps

Question and applicability

Inspect the matched virtual-tumour precision/recall evidence for Lancet and Strelka2 before selecting a caller and designing a validation protocol for the intended sequencing/normal regime.

Who this is for: Clinical researchers scoping the available diagnostic-genomics evidence
Computational researchers comparing exact evaluated configurations
Research setting: Research and clinical research. See Clinical research scope for what the evidence does not establish.
Biological setting: A real-read virtual-tumour spike-in (NA12892/NA12891 HapMap samples, 80x/40x WGS) measures SNV and indel precision/recall/F1 for the Lancet caller and the Strelka2 comparator.

Outside this use case

Real clinical tumour specimens, FFPE samples, targeted panels, ctDNA, copy-number and structural variation
Callers other than Lancet/Strelka2 not transcribed in this bounded intake (Strelka, MuTect, MuTect2, LoFreq rows exist in the source but are not all ingested; LoFreq/MuTect2 unselected rows are internally inconsistent in the source and are excluded)

Inputs and expected output

Inputs you need

Tumour and matched-normal aligned sequencing reads
The intended variant-fraction, coverage and genomic-context strata

Expected output

A sourced set of exact evaluated caller precision/recall/F1 values and remaining validation needs; no patient-level variant classification or clinical recommendation.

Clinical research scope

Clinical applicability is not established. A synthetic virtual-tumour spike-in does not demonstrate sensitivity in real clinical tumours, FFPE specimens, gene panels, ctDNA, or for copy-number/structural variants.

Evaluated evidence

Evidence is grouped by its protocol. Relevance refers to the stated endpoint and context; it is separate from clinical validation and from the review method. Limits specific to each evaluation are listed with it.

Lancet paper virtual-tumor SNV/indel benchmark

Current source-reviewed mapping

Direct evidence for the stated endpoint

Real-read virtual-tumour spike-in directly measures somatic SNV/indel calling precision and recall, the declared benchmark endpoint, under one sequencing/normal regime.

Assessed endpoint

Virtual-tumour SNV and indel precision/recall/F1, true/false positive and false negative counts for Lancet and Strelka2 on a real-read synthetic tumour/normal pair

Evaluation protocol

Lancet paper virtual-tumor SNV/indel benchmark

Computational task

A reviewed task relationship is not recorded for this protocol.

Input and population constraints

Inspect every linked evaluation's source locator and preserved conflicts before citing a result.
Do not combine this mapping's evaluations with any other protocol's results.

Limits on interpretation

Author-developed-caller evaluation, not independent reproduction.
Virtual tumor endpoint does not establish sensitivity in clinical tumors, FFPE samples, panels, ctDNA, copy-number or structural variation.
SNV printed counts agree with 31,592 truth variants. Do not repair MuTect table #calls=50,228 versus TP+FP=26,505; its row is not proposed for numerical ingestion.
Indel Lancet and Strelka2 counts agree with 4,945 truth variants. Preserve Table 1 rounding and do not recompute percentages.
Individual-condition confidence intervals unreported in Tables 1–2.

Automated source review · 2026-10-07 · Claude Sonnet AMP-integration worker, bounded transcription of Codex-checked primary values; independently reviewed by Codex (workbench/amp-supervision/primary-review.md, integration-review-corrections.md)

Bounded primary-source transcription, independently Codex-checked. No new model execution, independent experimental reproduction, qualified human scientific review or clinical validation.

Evaluated configurations

Each configuration below belongs to this protocol. Inspect its inputs, population and scoring conditions before comparing it with another evaluation.

Lancet

Author-reported evaluation · Source checked

Lancet evaluation; bounded primary-source AMP candidate.

Inspect results, conditions and reproduction (12 recorded results)

Population and split: 31,592 SNVs / 4,945 indels · Not reported
Inputs and adaptation: virtual-tumour and matched-normal aligned BAM · Not reported
Evaluation budget: Not reported
Runtime and memory: Runtime and memory measurements are not reported in this evaluation. A study budget is not a runtime or memory measurement.

Recorded results for this configuration
Metric	Value	Coverage	Uncertainty and source
indel F1-score	0.81 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Lancet row (JATS Tab1)`
indel false negatives	1360 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Lancet row (JATS Tab1)`
indel false positives	305 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Lancet row (JATS Tab1)`
indel precision	0.92 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Lancet row (JATS Tab1)`
indel recall	0.72 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Lancet row (JATS Tab1)`
indel true positives	3590 variants · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Lancet row (JATS Tab1)`
SNV F1-score	0.85 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Lancet row (JATS Tab2)`
SNV false negatives	7740 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Lancet row (JATS Tab2)`
SNV false positives	565 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Lancet row (JATS Tab2)`
SNV precision	0.98 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Lancet row (JATS Tab2)`
SNV recall	0.75 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Lancet row (JATS Tab2)`
SNV true positives	23,848 variants · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Lancet row (JATS Tab2)`

Evaluation methods, evidence and reproduction

No execution recipe has been verified for this exact configuration and evaluation. Inspect its methods and original run documentation before attempting reproduction.

Strelka2

Independent external evaluation · Source checked

Strelka2 evaluation; bounded primary-source AMP candidate.

Inspect results, conditions and reproduction (12 recorded results)

Population and split: 31,592 SNVs / 4,945 indels · Not reported
Inputs and adaptation: virtual-tumour and matched-normal aligned BAM · Not reported
Evaluation budget: Not reported
Runtime and memory: Runtime and memory measurements are not reported in this evaluation. A study budget is not a runtime or memory measurement.

Recorded results for this configuration
Metric	Value	Coverage	Uncertainty and source
indel F1-score	0.77 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Strelka2 row (JATS Tab1)`
indel false negatives	1300 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Strelka2 row (JATS Tab1)`
indel false positives	867 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Strelka2 row (JATS Tab1)`
indel precision	0.81 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Strelka2 row (JATS Tab1)`
indel recall	0.73 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Strelka2 row (JATS Tab1)`
indel true positives	3650 variants · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 1, Strelka2 row (JATS Tab1)`
SNV F1-score	0.85 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Strelka2 row (JATS Tab2)`
SNV false negatives	7460 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Strelka2 row (JATS Tab2)`
SNV false positives	1120 variants · lower	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Strelka2 row (JATS Tab2)`
SNV precision	0.96 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Strelka2 row (JATS Tab2)`
SNV recall	0.76 fraction · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Strelka2 row (JATS Tab2)`
SNV true positives	24,132 variants · higher	Not reported scored / Not reported eligible	Not reported Result provenance Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗`Table 2, Strelka2 row (JATS Tab2)`

Evaluation methods, evidence and reproduction

No execution recipe has been verified for this exact configuration and evaluation. Inspect its methods and original run documentation before attempting reproduction.

Open the protocol's results and comparison checks →

Mapping sources and review metadata

Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗See clinical/claims.csv and clinical/sources.md in data/omics/use-case-coverage-amp-20261007/

Mapping use-case-mapping-amp-20261007-issue10-lancet-virtual-tumor · revision 1

Add Codex-checked primary-source protocol evidence from the bounded AMP intake (rewire.it#365).

Reviewed evidence fingerprint 018e83af740b165027737c9b171ceb64ed144c548de9666ac12367c215027b78

Limitations and missing evidence

These gaps apply to the question as a whole. Absence of evidence is not a zero score.

Individual-condition confidence intervals are unreported in the source tables.
Exact caller release versions are not extracted from the primary article.
No foundation-model component is evaluated in this bounded intake.

Planned work

These plans do not contribute measured results or evaluated winners above.

Focused benchmark/run task for rewire-benchmark-data issue #10 ↗

Planned · Reviewed literature evidence is bounded (see evidence_gaps); closing the remaining decision gap needs a dedicated protocol/run task with frozen population and matched controls, tracked in rewire-benchmarks.

Contribute evidence or propose a correction

Evidence collection plan

Collecting evidence

Mapped evidence already covers 1 evaluated endpoint, in the evaluated evidence above. The status above describes only the specific comparison in this plan, which remains open; it does not mean no evidence has been collected. The plan defines a comparison to investigate; it does not establish model performance or suitability.

Comparison question

Which calling or rescoring workflow reliably detects tumour SNVs and small indels under the available sequencing and normal-sample regime?

Baselines, outcomes and validation requirements

Baselines to include

The conventional/author-introduced workflow measured in the linked primary source(s).

Outcomes to measure

The declared endpoint in this use case's active mapping(s); see evidence_gaps for what remains open.

Validation requirements

Independent held-out population matched to the intended clinical setting.
Qualified human scientific review before any clinical-validation claim.

Next collection task

A dedicated rewire-benchmarks protocol/run task with a real clinical tumour cohort, pinned truth set, purity/platform/assembly strata and an explicit foundation-model eligibility check.

Sources and review

Bounded primary-source transcription, independently Codex-checked. No new model execution, independent experimental reproduction, qualified human scientific review or clinical validation.

Genome-wide somatic variant calling using localized colored de Bruijn graphs · Original source ↗See data/omics/use-case-coverage-amp-20261007/clinical/sources.md

Release provenance and downloads

Release 2026-10-07-1448159e6a81

Use-case input digest d60fd7f669bfec7bd34ec6e5080d8e1cb4e8186f8286ead60888848f1e20e001

Download questions, collection plans and review metadata (JSON) · Verify release checksums

Question use-case-tumour-dna-somatic-variant-detection. Any numerical results on this page come from this release's existing evaluation records.