Interpreting genomic variants is one of the most time-consuming and complex tasks in clinical genomics. Each variant must be assessed for its clinical significance using established guidelines (where available), curated databases, and the latest published evidence. Manual review can be slow and inconsistent, particularly as new findings continually reshape the landscape. 

VarSome Clinical supports this process with two automated classification engines – germline and somatic. Both are designed to deliver transparent, evidence-based results that save time while preserving expert oversight. 

Germline Variant Classification

The VarSome Germline Classifier is based on the ACMG guidelines to deliver pathogenicity recommendations: Pathogenic, Likely Pathogenic, Variant of Uncertain Significance, Likely Benign, and Benign. The classifier is powered by over 140 data sources, including gnomAD, ClinVar, other clinical databases, in silico predictors, conservation data, and published studies. 

Key features:

• ACMG Rule Logic: Implements 21 rules with context-aware strength adjustments, ensuring consistent application of evidence.
• Splicing-aware Interpretation: Incorporates predictors like scSNV and MaxEntScan to identify variants affecting splice sites.
• Transparency: Every triggered and untriggered rule is fully documented in plain language, with links to supporting data. In the germline classifier, you can also see how each rule contributes to the final score, which is calculated by weighting supporting and opposing evidence in line with ACMG guidelines.
• User Control: Clinicians can override rule strengths, select relevant transcripts, and adjust evidence weighting in real time. 

For laboratories, this means consistent and reproducible germline classifications that can be integrated into variant discovery, annotations, and reporting pipelines.

To read more about our germline classifier, please visit VarSome Germline Classification.

Somatic Variant Classification

In oncology, variant interpretation requires a different approach, focusing on tumor-specific mutations and their clinical implications. The VarSome Somatic Variant Classifier, based on AMP/ASCO/CAP guidelines, classifies variants into tiers based on their therapeutic, diagnostic, or prognostic significance. 

Key Features:

• Tiered Classification: Assigns variants to categories such as strong clinical evidence, potential clinical significance, or benign/likely benign. 
• Clinical Relevance: Integrates curated therapeutic, diagnostic, and prognostic evidence, such as OncoKB, COSMiC, and JaxCKB, to support precision oncology. 
• Contextual Interpretation: Considers tumor type, literature evidence, and database insights for accurate classification. 
• Transparency and Traceability: As with germline classification, users see exactly which criteria were applied and why.

Our somatic classifier supports oncology teams in identifying actionable mutations, streamlining clinical decision-making, and aligning with best-practice frameworks. 

For more information on our somatic classifier, please see VarSome Somatic Classification.

Shared Principles

While the two classifiers serve different pipelines, they share common principles:

• Transparency: Every classification is explained with accessible evidence. 
• Configurability: Users can tailor settings to their laboratory's workflow.
• Consistency: The same logic applies across users and over time.
• Continuous Improvement: Both classifiers are updated regularly with new data, curated evidence, and community feedback.

Together, they enable laboratories to approach both hereditary and tumor variant interpretation with a single IVDR-certified platform, supporting fast decisions and more reliable outcomes. Laboratories can also reannotate variants at any time, ensuring analyses remain aligned with the most up-to-date data and guidelines. 

Conclusion

By combining automated rule evaluation with user configurability and a commitment to transparency, VarSome Clinical helps bring clarity and consistency to variant classification. Whether applied to germline or somatic variants, these classifiers help reduce the manual burden while supporting evidence-based, auditable decision-making. 

To read about each classifier in more detail, visit our Help Center pages VarSome Germline Classification and  VarSome Somatic Classification, or contact our team to see them in action.