Identifying genetic biomarkers associated to Alzheimer's disease using Support Vector Machine

Machine learning methods are used to identify genetic biomarkers associated to complex diseases. Alzheimer's disease (AD) is a degenerative disorder that attacks the brain's neurons. Single Nucleotide Polymorphisms (SNPs) are the most common type of human genetic variation. SNPs are useful markers for disease genes. SNPs related to many common and serious diseases like AD. Discovering SNP biomarkers associated with AD can contribute to early prediction and diagnosis of this disease. Feature selection methods namely Correlation-based feature selection (CFS) and Chi-squared feature selection were used to find the most important SNPs. Support Vector Machine (SVM) classifier of different kernels has been applied on Alzheimer's Disease Neuroimaging Initiative Phase 1 (ADNI-1) data based on selected 21 variants most associated with AD. Results revealed that SVM trained model using RBF kernel has a high association with AD and achieves better accuracy of 76.70%.