AI system
- System
-
Exomiser phenotype-driven variant prioritization (machine-learning-assisted scoring and ranking)
- Model
-
Not assessed
- Release
-
Not assessed
- Application developer
-
Not named in the cited sources; the pilot's phenotype-based prioritization analysis was led by Damian Smedley's team at Queen Mary University of London and Genomics England.
- Role
-
Not assessed
- Configuration
-
Not assessed
- Inputs and outputs
-
Not assessed
- Autonomy
-
Not assessed
- Workflow
-
Genome sequencing with family data across a national pilot. Structured phenotype terms of adequate quality. Exomiser ranking used alongside applied gene panels, with clinical review before reporting.
- Human oversight
-
Not assessed
- Relation to AI
-
Exomiser scored and ranked candidate variants against each proband's phenotype. The pilot reports that diagnoses were detected in 77% of top-ranked candidate variants and that panels and Exomiser's top five were complementary. Clinical confirmation and reporting were performed by NHS diagnostic laboratories and clinicians, not by the software. Neither cited source says who developed Exomiser. The pilot's prioritization analysis was led by Damian Smedley's team at Queen Mary University of London and Genomics England, who are among the paper's authors. An independent evaluation is cited alongside it.
- AI contribution
-
Exomiser scored and ranked candidate variants against each proband's phenotype. The pilot reports that diagnoses were detected in 77% of top-ranked candidate variants and that panels and Exomiser's top five were complementary. Clinical confirmation and reporting were performed by NHS diagnostic laboratories and clinicians, not by the software. Neither cited source says who developed Exomiser. The pilot's prioritization analysis was led by Damian Smedley's team at Queen Mary University of London and Genomics England, who are among the paper's authors. An independent evaluation is cited alongside it.
Benefits
- Outcome identifiers
-
Not assessed
- Benefit
-
Genetic diagnoses were found in 25% of probands, rising to 35% for monogenic disorders
- Category
-
Improved Physical Health
- Status
-
Not assessed
- Beneficiary groups
-
People. Organizations. Society.
- Beneficiaries
-
Not assessed
- Reach
-
Not assessed
- Distribution and access
-
Not assessed
- Duration
-
Not assessed
- Expected?
-
Yes
- Costs and harms
-
Not assessed
Measurement
- Studies
-
Not assessed
- Study design
-
Not assessed
- Measurement
-
Not assessed
- Setting and sample
-
Not assessed
- Measures
-
Genetic diagnoses were found in 25% of probands, rising to 35% for monogenic disorders
- Comparison
-
the virtual panel-based pipeline (applied gene panels) alone
- Results
-
Reference: A total of 322 (66%) of the 490 diagnoses that were based on SNVs or indels from the genomes were made with the virtual panel–based pipeline. Result: diagnoses were detected in 77% of the top-ranked candidate variants, in 86% of the top three candidates, and in 88% of the top five candidates.
- Additional value from AI
-
Not assessed
- Limitations
-
Not assessed
- Denominators
-
Not assessed
- Materials for replication
-
Not assessed
Money
- Kind of value
-
Not assessed
- Amount
-
Not assessed
- Currency and prices
-
Not assessed
- Period and unit
-
Not assessed
- Realization
-
Not assessed
- Recipients and payers
-
Not assessed
- Gross or net
-
Not assessed
- Included costs
-
Not assessed
- Comparison
-
Not assessed
- Method
-
Not assessed
- Return and payback
-
Not assessed
People and places
- Participants
-
Developer: Not named in the cited sources; the pilot's phenotype-based prioritization analysis was led by Damian Smedley's team at Queen Mary University of London and Genomics England. Operator: Genomics England and NHS Genomic Medicine Centers.
- Jurisdictions
-
Benefit: England,
United Kingdom. Operator: United Kingdom. - Ownership and control
-
Not assessed
- Dependencies
-
Not assessed
- Cross-border connections
-
Not assessed
- Geopolitical implications
-
Not assessed
Follow-up
- Conditions for replication
-
Genome or exome sequencing with family data where available. Structured HPO phenotype terms of adequate quality. Clinical review of ranked candidates before reporting.
- Replication attempts
-
1 follow-up entries. Benefit recurred: 1.
- Later findings
-
Not assessed
- Availability
-
Not assessed
- Institutional decision
-
Not assessed
- Grounds
-
Not assessed
- Action and result
-
Not assessed
- Success criterion
-
Not assessed
Evidence
- Source numbers
-
AIBD-0049-source-0001–0002
- Citations
-
Numbered references below.
- Language
-
Not assessed
- Source relationships
-
Operator, Independent
- Original evidence
-
Not assessed
- Access and versions
-
Not assessed
- Supported findings
-
Not assessed
- Evidence links
-
Not assessed
- Evidence strength
-
Not assessed
- Missing evidence
-
Not assessed
Grades
- Assessed benefits
-
People. Organizations. Society.
- Depth
-
Not assessed
- Reasons
-
People: Clinicians reported that 134 of 533 genetic diagnoses (25%) were of immediate clinical actionability, including medication changes and surveillance. That is a cohort figure and no individually attributable changed outcome is cited. Organizations: Exomiser's top five candidates together with applied panels accounted for 92% of the 490 diagnoses. That is an aggregate reduction in manual interpretation burden across the pilot rather than a discrete organizational event. Society: A 25% diagnostic yield across 2,183 families is reported for a national health-system pilot rather than population-wide deployment.
- Observed or intended
-
Not assessed
- Rubric
-
Not assessed
- Assessor and date
-
Not assessed
- Alternative assessments
-
Not assessed
- Fill
-
Independently studied
Entry details
- Kind
-
Not assessed
- Related entries
-
Both are machine-learning diagnostic decision-support models relied on across health systems; the COVID detection and prognostication models proved unfit for clinical use, harming COVID patients and their healthcare providers.
- Field
-
Clinical genomics / rare disease diagnosis. Families with undiagnosed rare disease in a national health system.
- Purpose
-
Score and rank candidate causative variants against patient phenotypes to make genome-scale rare-disease diagnosis tractable for clinical teams.
- Dates
-
Date basis unassessed: 2021 (precision unassessed). Updated: 21 October 2026.
- Stage
-
Not assessed
- Version
-
SHA-256: a1b860f3f4af…
References
- Other identifiers
-
AI Incident Database: 535.
- Category mappings
-
Not assessed
- Brief
-
Not assessed
- Revisions
-
7 revisions. Latest: 21 October 2026.
Sources
-
Peer-reviewed study Deployer
100,000 Genomes Project Pilot Investigators; Smedley D, Smith KR, Martin A, et al. (2021). 100,000 Genomes Pilot on Rare-Disease Diagnosis in Health Care - Preliminary Report. N Engl J Med.
With the use of phenotype-based prioritization with the Exomiser to score and rank the most likely causative variants, diagnoses were detected in 77% of the top-ranked candidate variants.
-
Peer-reviewed study Independent author
Cooperstein IB, Marwaha S, Ward A, et al. (2025). An optimized variant prioritization process for rare disease diagnostics: recommendations for Exomiser and Genomiser. Genome Med.
the percentage of coding diagnostic variants ranked within the top 10 candidates increased from 49.7% to 85.5%