novoCaller

novoCaller calls de novo variants using a Bayesian network to analyze read-level sequence data from pedigrees and unrelated samples, identifying newly occurring mutations relevant to sporadic dominant monogenic diseases and complex disorders.

Key Features:

• Bayesian Network Approach: Models the probability of de novo variants with a Bayesian network using read-level sequence data from pedigrees and unrelated individuals.
• High Sensitivity and Specificity: Demonstrated over 97% sensitivity on large trio sequencing studies and higher specificity than established methods at comparable sensitivity levels.
• Application in Genetic Research: Applied to 48 trios of suspected rare Mendelian disorders in the Brigham Genomic Medicine gene discovery initiative.
• Impact on Genetic Diagnosis: Identified three de novo variants in known disease-associated genes and discovered 14 novel variants in genes likely linked to the observed phenotypes.
• Resource Efficiency: Reduced the need for extensive manual inspection and experimental validation in trio analyses.

Scientific Applications:

• Monogenic Disease Research: Identification of de novo mutations to support diagnosis of sporadic dominant monogenic diseases.
• Complex Disorder Genetics: Prioritization of novel mutation candidates to inform studies of complex disorders.
• Population Genetics Models: Informing population genetics and studies of DNA replication and repair mechanisms.

Methodology:

Uses a Bayesian network to analyze read-level sequence data from pedigrees and unrelated samples.