Configuration

Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)

Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026) as run in Kaschta et al. 2026.

3 evaluations · 17 results

Overview

Talos: Inheritance-aware genomic variant prioritization and iterative reanalysis.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

Evaluations and results

3 evaluations · 17 results. Different protocols are not a single leaderboard.

Sorted by Candidates per case (average variants returned per case, singleton and trio cases) (lower is better). The best value in each column is highlighted. Decimals are rounded for display; each value links to the printed value and its source.

Tested configuration	Protocol and dataset	Candidates per case (average variants returned per case, singleton and trio cases)	Count (cases with a new P/LP finding)	Count (cohort cases with a reported P/LP finding after reanalysis (158 before))	Count (cohort cases without a prioritised finding after reanalysis (170 before))	Count (of the two cases newly classified as VUS by manual reanalysis, cases also prioritised)	Proportion (known P/LP variants missed because of HPO annotation gaps)	Proportion (known P/LP variants missed because of pipeline conversion errors)	Proportion (known P/LP variants missed: HPO annotation gaps)	Proportion (known P/LP variants missed: large copy-number variants not processed)	Proportion (known P/LP variants missed: out-of-scope variant classes)	Proportion (known P/LP variants missed: pipeline conversion errors)	Proportion (known P/LP variants not prioritised because of trio inheritance filtering)	Recall (known P/LP variants prioritised, trio mode)	Recall (known P/LP variants prioritised)	Count (cohort cases with VUS findings after reanalysis (49 before))	Count (known VUS prioritised, trio mode)	Count (known VUS prioritised)
Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)Evaluation	Manual versus Talos reanalysis of 219 unresolved genomes after a mean 660 days (Kaschta et al. 2026), UKSH rare-disease genomes without a prior P/LP finding: 219 cases (Kaschta et al. 2026)	three	three	161	167	one	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	49	Not reported	Not reported
Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)Evaluation	Talos recovery of known P/LP and VUS variants in 45 singleton cases (Kaschta et al. 2026), UKSH singleton benchmarking cases with known findings (Kaschta et al. 2026)	Not reported	Not reported	Not reported	Not reported	Not reported	11.4%	8.6%	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	80%	Not reported	Not reported	two
Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)Evaluation	Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026), UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	6.2%	2.1%	2.8%	6.2%	7.6%	75.2%	Not reported	Not reported	Six	Not reported

All 17 result rows with coverage, uncertainty and sources

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 45 singleton cases (Kaschta et al. 2026) Dataset: UKSH singleton benchmarking cases with known findings (Kaschta et al. 2026)	8.6% Proportion (known P/LP variants missed because of pipeline conversion errors) percent · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos on 45 singleton benchmarking cases reanalysis-kaschta-20261010-protocol-talos-singleton-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Singleton Cases P3, 'pipeline conversion errors (8.6%)'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 45 singleton cases (Kaschta et al. 2026) Dataset: UKSH singleton benchmarking cases with known findings (Kaschta et al. 2026)	11.4% Proportion (known P/LP variants missed because of HPO annotation gaps) percent · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos on 45 singleton benchmarking cases reanalysis-kaschta-20261010-protocol-talos-singleton-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Singleton Cases P3, 'HPO annotation gaps (11.4%)'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 45 singleton cases (Kaschta et al. 2026) Dataset: UKSH singleton benchmarking cases with known findings (Kaschta et al. 2026)	80% Recall (known P/LP variants prioritised) percent · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos on 45 singleton benchmarking cases reanalysis-kaschta-20261010-protocol-talos-singleton-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Singleton Cases P2, 'Talos prioritized 28 out of 35 P/LP variants (80.0% concordance)'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 45 singleton cases (Kaschta et al. 2026) Dataset: UKSH singleton benchmarking cases with known findings (Kaschta et al. 2026)	two Count (known VUS prioritised) count · unknown Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos on 45 singleton benchmarking cases reanalysis-kaschta-20261010-protocol-talos-singleton-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Singleton Cases P2, 'only two out of 21 VUS were captured'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Manual versus Talos reanalysis of 219 unresolved genomes after a mean 660 days (Kaschta et al. 2026) Dataset: UKSH rare-disease genomes without a prior P/LP finding: 219 cases (Kaschta et al. 2026)	three Candidates per case (average variants returned per case, singleton and trio cases) variants-per-proband · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos reanalysis of 219 unresolved genomes reanalysis-kaschta-20261010-protocol-manual-vs-talos-219-reanalysis Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Reanalysis Diagnostic Yield: Automated Reanalysis P1, 'On average, Talos returned three variants per case'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Manual versus Talos reanalysis of 219 unresolved genomes after a mean 660 days (Kaschta et al. 2026) Dataset: UKSH rare-disease genomes without a prior P/LP finding: 219 cases (Kaschta et al. 2026)	three Count (cases with a new P/LP finding) count · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos reanalysis of 219 unresolved genomes reanalysis-kaschta-20261010-protocol-manual-vs-talos-219-reanalysis Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Reanalysis Diagnostic Yield: Automated Reanalysis P1, 'also identified the three new P/LP cases'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Manual versus Talos reanalysis of 219 unresolved genomes after a mean 660 days (Kaschta et al. 2026) Dataset: UKSH rare-disease genomes without a prior P/LP finding: 219 cases (Kaschta et al. 2026)	one Count (of the two cases newly classified as VUS by manual reanalysis, cases also prioritised) count · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos reanalysis of 219 unresolved genomes reanalysis-kaschta-20261010-protocol-manual-vs-talos-219-reanalysis Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Abstract P3, 'only identified one of the two new VUS findings'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Manual versus Talos reanalysis of 219 unresolved genomes after a mean 660 days (Kaschta et al. 2026) Dataset: UKSH rare-disease genomes without a prior P/LP finding: 219 cases (Kaschta et al. 2026)	167 Count (cohort cases without a prioritised finding after reanalysis (170 before)) count · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos reanalysis of 219 unresolved genomes reanalysis-kaschta-20261010-protocol-manual-vs-talos-219-reanalysis Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Reanalysis Diagnostic Yield: Automated Reanalysis P1, 'decreased from 170 to 167'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Manual versus Talos reanalysis of 219 unresolved genomes after a mean 660 days (Kaschta et al. 2026) Dataset: UKSH rare-disease genomes without a prior P/LP finding: 219 cases (Kaschta et al. 2026)	161 Count (cohort cases with a reported P/LP finding after reanalysis (158 before)) count · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos reanalysis of 219 unresolved genomes reanalysis-kaschta-20261010-protocol-manual-vs-talos-219-reanalysis Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Reanalysis Diagnostic Yield: Automated Reanalysis P1, 'increased the number of cases with a P/LP finding from 158 to 161'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Manual versus Talos reanalysis of 219 unresolved genomes after a mean 660 days (Kaschta et al. 2026) Dataset: UKSH rare-disease genomes without a prior P/LP finding: 219 cases (Kaschta et al. 2026)	49 Count (cohort cases with VUS findings after reanalysis (49 before)) count · unknown Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos reanalysis of 219 unresolved genomes reanalysis-kaschta-20261010-protocol-manual-vs-talos-219-reanalysis Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Reanalysis Diagnostic Yield: Automated Reanalysis P1, 'The number of cases based on results with VUS findings remained at 49'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026) Dataset: UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	6.2% Proportion (known P/LP variants missed: pipeline conversion errors) percent · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos in trio mode on 162 trio benchmarking cases reanalysis-kaschta-20261010-protocol-talos-trio-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Trio Cases P1, 'nine pipeline conversion errors (6.2%)'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026) Dataset: UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	6.2% Proportion (known P/LP variants missed: HPO annotation gaps) percent · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos in trio mode on 162 trio benchmarking cases reanalysis-kaschta-20261010-protocol-talos-trio-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Trio Cases P1, 'nine HPO annotation gaps (6.2%)'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026) Dataset: UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	7.6% Proportion (known P/LP variants not prioritised because of trio inheritance filtering) percent · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos in trio mode on 162 trio benchmarking cases reanalysis-kaschta-20261010-protocol-talos-trio-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Trio Cases P2, 'An additional 11 P/LP variants (7.6%) were not prioritized'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026) Dataset: UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	2.1% Proportion (known P/LP variants missed: large copy-number variants not processed) percent · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos in trio mode on 162 trio benchmarking cases reanalysis-kaschta-20261010-protocol-talos-trio-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Trio Cases P1, 'three large copy-number variants not processed by the pipeline (2.1%)'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026) Dataset: UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	2.8% Proportion (known P/LP variants missed: out-of-scope variant classes) percent · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos in trio mode on 162 trio benchmarking cases reanalysis-kaschta-20261010-protocol-talos-trio-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Trio Cases P1, 'four out-of-scope variant classes (2.8%)'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026) Dataset: UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	75.2% Recall (known P/LP variants prioritised, trio mode) percent · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos in trio mode on 162 trio benchmarking cases reanalysis-kaschta-20261010-protocol-talos-trio-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Trio Cases P1, 'overall P/LP concordance of 75.2%'
Configuration: Talos 8.2.0 on archived DRAGEN v3.7.5 VCFs, pedigree where available (Kaschta et al. 2026)	Protocol: Talos recovery of known P/LP and VUS variants in 162 trio cases (Kaschta et al. 2026) Dataset: UKSH trio benchmarking cases with known findings (Kaschta et al. 2026)	Six Count (known VUS prioritised, trio mode) count · unknown Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Talos in trio mode on 162 trio benchmarking cases reanalysis-kaschta-20261010-protocol-talos-trio-benchmark Aggregation: Not reported Automated versus manual reanalysis in rare disease genomics · Results, Benchmarking of Automated Reanalysis: Trio Cases P1, 'Six VUS variants were captured'

Source checking is not independent reproduction. Release 2026-10-10-cbb3da59bc08.

Use this configuration

How it works, versions and access

Strengths, limitations and unresolved questions

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

0 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-10-10-cbb3da59bc08
Property and statement	Original source and location	Review and provenance

No evidence rows match these filters. Choose another scope or clear the search.

All evidence (JSONL) (gzip)How to read the evidence

Sources and history

Release 2026-10-10-cbb3da59bc08 · Record review: source checked

1 source record and release history

Automated versus manual reanalysis in rare disease genomics · Original source · medRxiv 2026.05.16.26352295 v1, posted 2026-05-19; JATS source XML

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: reanalysis-kaschta-20261010-config-talos-8-2-0-archived-dragen-vcf

areas: dna-genomes
contexts: clinical_research
method types: conventional_pipeline
reported name: Talos
foundation model eligible: false
version: 8.2.0
parameters: Workflow customised to accept VCFs already annotated by DRAGEN; archived per-case VCFs, pedigree information where available and optional HPO terms; no manual curation; trio mode applies inheritance filtering
source locator: Methods 'Study Cohort and Reanalysis Design' P3; 'Automated Reanalysis Using Talos' P1-P4
limitations: Adapted by the authors to ingest DRAGEN-annotated VCFs; conversion errors stopped some variants entering prioritisation (Results).

Related records

configuration of: Talos
system: Talos reanalysis of 219 unresolved genomes
system: Talos on 45 singleton benchmarking cases
system: Talos in trio mode on 162 trio benchmarking cases

Suggest a correction