Alz-disc

Alz-disc classifies mutations associated with Alzheimer’s disease as disease-causing or neutral using sequence-derived features and a Bayes network-based machine learning algorithm.

Key Features:

• Curated Mutation Dataset: Utilizes a dataset of 314 Alzheimer’s disease-causing mutations and 370 neutral mutations for model training and validation.
• Sequence Feature Extraction: Computes sequence-based features including conservation scores, position-specific scoring matrix (PSSM) profiles, hydrophobicity changes, amino acid substitution matrices, and neighboring residue information.
• Bayesian Network Classification: Applies a Bayes network-based algorithm to model relationships among sequence features and classify mutations.

Scientific Applications:

• Variant Pathogenicity Prediction: Predicts whether genetic variants are likely to be disease-causing or neutral in Alzheimer’s disease.
• Genetic Variant Annotation: Supports functional annotation of newly identified mutations in Alzheimer’s disease research.
• Neurogenetic Research: Assists studies investigating the molecular and genetic mechanisms underlying Alzheimer’s disease.

Methodology:

Alz-disc extracts sequence-derived features such as conservation scores, PSSM profiles, hydrophobicity changes, substitution matrix values, and neighboring residue information and applies a Bayes network-based machine learning algorithm to classify mutations as disease-causing or neutral.