Evidence profiles
Profiles are reusable checklists of evidence rules. They define what should be verified, how each finding is interpreted, and whether accountable evidence is present during an analysis.
| Profile | Domain | Usage signal | Rules | Action |
|---|---|---|---|---|
Single proband WGS candidate variant review where evidence is assessed across genotype validity, population frequency, public clinical assertions, gene disease consistency, phenotype fit, control observations, and UniProt functional annotation. | Clinical Genetics | New0 projects · 0 cases · 0 runs | 21 | |
Single proband WGS candidate variant review where evidence is assessed across genotype validity, population frequency, public clinical assertions, gene disease consistency, phenotype fit, control observations, and UniProt functional annotation. | Clinical Genetics | New0 projects · 0 cases · 0 runs | 21 | |
Each evaluated item is an analysis run, model output, notebook, script, or response that claims to analyse Apple Inc. stock price data. The profile checks whether the analysis used the agreed externally verifiable source fields open, high, low, close, adjusted, and volume, and whether the analysis produced an audit manifest and checksums. A FALSE raw outcome means the required evidence is present and inspectable. A TRUE raw outcome means the required evidence is missing, inconsistent, inaccessible, or not inspectable. NA means the check could not be evaluated and contributes no evidence under SGA-QEM-1.0. | General | New0 projects · 0 cases · 0 runs | 20 | |
Candidate target review where each candidate must have accountable evidence across available omics layers. | Multiomics | New0 projects · 0 cases · 0 runs | 7 | |
Trio-based whole-genome sequencing for rare Mendelian disease screening | Clinical Genetics | New0 projects · 0 cases · 0 runs | 5 | |
Multiomics compound screening | Other | New0 projects · 0 cases · 0 runs | 4 | |
Wet lab plasmid and construct verification | Wet Lab | New0 projects · 0 cases · 0 runs | 4 |