CoDP

CoDP predicts the impact of missense variants in the MSH6 gene to assess their pathogenicity relevant to Lynch syndrome and DNA mismatch repair deficiency.

Key Features:

• Integration of prediction methods: Combines scores from MAPP, PolyPhen-2, and SIFT for variant impact assessment.
• Structural properties: Incorporates solvent accessibility and changes in the number of heavy atoms of amino acids within MSH6.
• Logistic regression model: Applies a logistic regression fitted with parameters derived from clinical and molecular data to predict pathogenicity.
• Performance metrics: Reports PPV 93.3%, NPV 94.7%, specificity 94.7%, sensitivity 93.3%, accuracy 94.1%, and AUC 0.954 compared with MAPP 0.919, SIFT 0.864, and PolyPhen-2 HumVar 0.819.
• Statistical validation: Distinguishes pathogenic and non-pathogenic variants with significance assessed by the Wilcoxon rank sum test (p < 8.9 × 10^(-6)).

Scientific Applications:

• Genetic testing for Lynch syndrome: Improves assessment of MSH6 missense variants to support variant classification in clinical testing.
• Variant interpretation for risk assessment: Aids prediction of pathogenicity to inform clinical risk assessment and management decisions.
• MSH6-focused research: Supports research into the molecular and clinical consequences of germline MSH6 missense variants.

Methodology:

Collected 294 germline missense variants from variant databases and literature, used a subset of 34 variants for parameter training and performance testing, combined MAPP, PolyPhen-2, and SIFT scores with solvent accessibility and amino-acid heavy-atom change features in a logistic regression framework, and evaluated performance by PPV, NPV, specificity, sensitivity, accuracy, AUC, and the Wilcoxon rank sum test.