Developing bioinformatics approaches to analyze and cluster pathogenic bacteria based on seg mental genomic duplication

Abstract

This project aims to implement the computer science concepts in the biotechnology field. The idea is to apply machine learning algorithms such as Fuzzy C-Means, Subtractive, and genetic algorithms. A set of pathogenic and non-pathogenic bacterium is selected to be clustered based on its genomic duplication features. The clustering is done by extracting a set of features from the genomic duplication in the DNA sequence of each bacterium. And then the correlation between the clusters and a group of biological features is calculated. To select the best combination of duplication features a genetic algorithm is used, each clustering process is evaluated and fitness is calculated, and the genetic algorithm select the best fitness. A hierarchical clustering is implemented on each of the duplication features, so we can analyze the feature from one dimension. The output of the hierarchical clustering is analyzed manually.