AUROC (all cancers vs non-cancer) of Fragmentia-AI WGS, high/medium/low-TF sequential fine-tuning (final model) on DELFI cohort from FinaleDB as used by the Fragmentia-AI WGS paper (538 samples)
0.929 AUROC (all cancers vs non-cancer)
Methods
- Tested configuration
- Fragmentia-AI WGS, high/medium/low-TF sequential fine-tuning (final model)
- Protocol
- Fragmentia-AI WGS paper: External DELFI cohort (FinaleDB)
- Dataset
- DELFI cohort from FinaleDB as used by the Fragmentia-AI WGS paper (538 samples)
- Procedure
- Fragmentia-AI WGS paper: External DELFI cohort (FinaleDB)
- Evaluation
- Fragmentia-AI WGS high/medium/low-TF sequential fine-tuning (final model) on the external delfi cohort (finaledb)
- Coverage
- Not reported scored / Not reported eligible
- Uncertainty
- 95% CI 0.907 to 0.951. Method not stated
- Evidence
- Author-reported evaluation · source checkedGeneralizable cancer detection from ultra-low-pass WGS via deep contextual modeling of cfDNA sequences; Fragmentia-AI WGS paper, Supplementary Material 1 (Tables S1-S4, Figures S1-S9) · Results P21, sentence 3
A source-checked result verifies the numerical transcription, not every model or protocol detail. Evaluation metadata: source checked. Source checked does not mean independently reproduced.
Reproduction
- Split
- External public cohort
- Adaptation
- high/medium/low-TF sequential fine-tuning (final model)
- Scoring implementation
- Not reported
No execution recipe has been verified for this exact configuration and evaluation. A benchmark's general instructions may use different inputs, splits or model settings.
Reproducing this published result requires matching its model configuration, data, split and scorer. Source checking or a successful smoke test does not establish score reproduction.
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.
2 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Reported result 0.929 Individual claims | Generalizable cancer detection from ultra-low-pass WGS via deep contextual modeling of cfDNA sequences Results P21, sentence 3 Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: Molecular Biomedicine, published online 2026-08-12; Europe PMC full-text JATS XML of PMC13469060 | source checked ["source-hash-verification","transcription","independent-cell-check"] · 2026-10-12 author reported Audit detailsRe-downloaded the pinned artifacts and matched their SHA-256; read Table S4 from the DOCX table XML with a separate reader written for this review (the extractor's script was not run) and checked each sensitivity against detected/total; compared the text values with Results P13, P14, P19 and P21. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record Extraction artifact SHA-256: |
| Reported result 0.929 Individual claims | Fragmentia-AI WGS paper, Supplementary Material 1 (Tables S1-S4, Figures S1-S9) Results P21, sentence 3 Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 43556_2026_538_MOESM1_ESM.docx, publisher supplementary file | source checked ["source-hash-verification","transcription","independent-cell-check"] · 2026-10-12 author reported Audit detailsRe-downloaded the pinned artifacts and matched their SHA-256; read Table S4 from the DOCX table XML with a separate reader written for this review (the extractor's script was not run) and checked each sensitivity against detected/total; compared the text values with Results P13, P14, P19 and P21. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record Extraction artifact SHA-256: |
Sources and history
Release 2026-11-21-c8bea25e5abb · Record review: source checked
2 source records and release history
- Generalizable cancer detection from ultra-low-pass WGS via deep contextual modeling of cfDNA sequences · Original source · Molecular Biomedicine, published online 2026-08-12; Europe PMC full-text JATS XML of PMC13469060
- Fragmentia-AI WGS paper, Supplementary Material 1 (Tables S1-S4, Figures S1-S9) · Original source · 43556_2026_538_MOESM1_ESM.docx, publisher supplementary file
Technical metadata and extraction receipts
Stable ID: ctdnalit-20261011-fragmentiawgs2026-result-delfi-high-medium-low-tf-auroc-overall
- metric
- auroc
- metric qualifier
- all cancers vs non-cancer
- metric direction
- higher
- unit
- unitless
- printed value
- 0.929
- numeric value
- 0.929
- source locator
- Results P21, sentence 3
- review
- method: source-hash-verification; transcription; independent-cell-check; reviewer: claude; reviewer note: Separate Claude review agent (reviewer-ctdna), independent of the extractor uc-ctdna; no human review claimed; date: 2026-10-12; artifact sha256: 8c2cd974ba60d7d29361adf31ee02bc705d7a5226b8423d47ecfdc32652b5ead; retrieval url: https://www.ebi.ac.uk/europepmc/webservices/rest/PMC13469060/fullTextXML; method note: Values parsed by regular expression from Results sentences asserted verbatim in extract/extract_fragmentiawgs2026.py.; note: Re-downloaded the pinned artifacts and matched their SHA-256; read Table S4 from the DOCX table XML with a separate reader written for this review (the extractor's script was not run) and checked each sensitivity against detected/total; compared the text values with Results P13, P14, P19 and P21.
- uncertainty
- type: confidence_interval; lower: 0.907; upper: 0.951; level: 0.95; printed: 95% CI: 0.907–0.951; note: Method not stated