Evidence page
The Evidence page presents the most relevant evidence supporting the automatic variant classification generated by the AI Shortlist. This includes ACMG classification tags, variant details, gene-disease relationships, phenotype matches, and supporting literature. You can also manually review and edit evidence to tailor it to your case.
The Evidence page is available for:
Variants in the AI Shortlist (automatically)
Any manually tagged variant in the analysis
Accessing the Evidence Page
You can view teh Evidence Page from:
Click on the variant bar to open its dedicated Evidence page.
In the Evidence section, click the See evidence button under the Evidence box.
View modes
You can switch between two display modes:
Graph view: Offers an interactive visual of gene-disease-phenotype relationships, inheritance modes, and ACMG tags.
Text view: Displays structured evidence and citations for use in clinical reports or further notes.
AI Shortlist collects data from credible studies and public databases in an internal knowledge base that maps complex connections between variants, genes, mechanisms, diseases, and phenotypes.
What's included in the Evidence Page?
Variant information
Displays the main effect, HGVS nomenclature, zygosity (HET, HOM, HEMI, or REF), and de novo status if applicable.
Gene information
Gene symbol, if the gene is tolerant to variation, and assumed inheritance mode in the case under review. This is suggested based on the observed level of genotype-phenotype co-segregation and inheritance mode of the genetic condition (reported or suspected).
In addition to the conventional inheritance modes (Autosomal Dominant, Autosomal Recessive, Compound Heterozygote Autosomal Recessive, X-Linked Dominant, X-Linked Recessive), the platform also employs Autosomal Dominant Partial Penetrance and Partial Autosomal Recessive designations.
Autosomal Dominant Partial Penetrance is used when the gene-associated condition is AD, and the variant is Het in the test subject and at least one of their parents. This suggests that the phenotypes may be due to incomplete penetrance of the genetic condition.
Partial Autosomal Recessive is used when the gene-associated condition is AR and the variant in the test subject is Het. This helps to account for the possibility that another causative variant is undetected - in the same (compound heterozygosity) or another (digenic inheritance) gene.
Disease information
Name of the condition as sourced from OMIM, literature, or other databases. When multiple diseases are associated with a gene, users can manually select the one most relevant to the case or choose to add a custom disease.
Patient phenotypes
Proband's phenotypes that match phenotypes reported for the suspected disease. Exact, indirect, and matches by ascendance are considered.
Unconfirmed disease phenotypes
Phenotypes reported for the suspected disease but not observed in the proband.
Unmatched patient phenotypes
Phenotypes observed in the proband but not known to be manifested as part of the suspected disease.
Links
Follow the links to the primary sources to explore the evidence further. The links are in the References section (text view) and are accessible by hovering over the arrows (graph view).
Edit mode
Click the pencil icon in the top-left corner of the graph to:
Edit disease associations
Add or remove evidence boxes
Tailor interpretation to case-specific insights
Important Notes When Editing the Evidence Graph
After modifying the evidence graph on the variant page or within the candidates section, the phenotypic match strength indicators may no longer appear in the side panel or on the variant page.
Changing the disease in the evidence graph does not automatically update the inheritance mode. If the inheritance pattern differs for the new disease, you must manually adjust the inheritance setting to ensure accurate interpretation.
Evidence Graph
This visual module links genes, diseases, and inheritance modes, providing an at-a-glance overview of your variant’s clinical context.
Easily switch the associated disease for reporting or interpretation.
The graph influences ACMG tags like PP4, PM5, and PVS1, depending on gene-disease pairings.
Connects with literature via LitVar2 integration.
Important Warning:
Changing the disease in the Evidence Graph does not automatically update inheritance mode. You must edit this manually to ensure consistency.
After editing, phenotypic match scores may disappear from the sidecar and Evidence Page.
Evidence Notes
Add your own insights or justifications related to the variant, interpretation, or evidence decisions.
Notes are tag-specific and can support audit trails and future reviews.
Note changes are not logged in the Activity Panel—consider logging important insights as Comments instead.
Activity tracking for Evidence
All meaningful updates—including ACMG tag edits, classification changes, and disease graph modifications—are logged in the Activity Panel, ensuring full transparency for collaborative curation and audits.
How to use the Evidence Page
Use Phenomeld scores (available as a filter and sortable column in Analysis Tools) to prioritize variants with high phenotype concordance.
When editing any ACMG tag, always click Save before editing notes to ensure no data is lost (especially for finalized cases).
Double-check that inheritance mode matches disease after editing the Evidence Graph.
For CNVs >20MB, perform manual evaluation, as automation and tagging will not apply.
The Evidence Page brings together automation, clinical reasoning, and interpretative flexibility, enabling users to make transparent, reproducible decisions in line with the latest guidelines—while giving you full control to adjust when needed.
Last updated
Was this helpful?