ClinPrior: an algorithm for diagnosis and novel gene discovery by network-based prioritization

Area of Science:

• Genomics
• Bioinformatics
• Rare Diseases

Background:

• Whole-exome sequencing (WES) and whole-genome sequencing (WGS) are crucial for diagnosing rare Mendelian genetic conditions.
• Existing diagnostic algorithms often struggle with incomplete gene-phenotype data and lack real-world validation.
• There is a need for faster, more sensitive algorithms to increase the diagnostic yield of WES/WGS in rare disease patients.

Purpose of the Study:

• To develop and validate ClinPrior, a novel algorithm for prioritizing candidate causal variants from WES/WGS data.
• To improve the diagnostic yield for rare genetic conditions by integrating patient phenotype information with interactome network analysis.
• To identify novel disease-associated genes and enhance the understanding of hereditary spastic paraplegia (HSP) and cerebellar ataxia (CA).

Main Methods:

• Developed ClinPrior, an algorithm that ranks candidate variants using standardized phenotypic features (Human Phenotype Ontology terms).
• Employed an interactome network-based approach for data propagation and variant prioritization.
• Benchmarked ClinPrior on a synthetic cohort and tested it on 135 families with HSP and/or CA.

Main Results:

• ClinPrior achieved a 70% positive diagnostic yield in a real-world cohort of rare disease patients.
• Identified 10 novel candidate genes associated with rare diseases, with 7 functionally validated.
• Generated a specific interactome for HSP/CA disorders, facilitating future diagnoses and gene discovery.

Conclusions:

• ClinPrior effectively improves clinical genomic diagnosis by integrating phenotype and interactome data.
• The algorithm aids in identifying atypical disease presentations and predicting novel disease-causing genes.
• ClinPrior contributes to increasing diagnostic yield, shortening diagnostic odysseys, and advancing the understanding of human genetic illnesses.