The main objective of the course consists in providing methodological tools for bioinformatics and computational biology, in order to analyze complex biomolecular data through machine learning methods.
0. Introduction to the R language for Bioinformatics (preliminary module of the course). 1. Bioinformatics as a multi-disciplinary approach to "omics" disciplines (genomics, proteomics, transcriptomics). 2. Non supervised methods. 2.1 Clustering algorithms for the analysis of omics data: k-means, fuzzy k-means, hierarchical clustering, self-organizing maps 2.2 Stability-based analysus of the reliability of the discovered clusters. Application to the discovery of pathological classes of patients. 3. Supervised methods. 3.1 Functional enrichment analysis. Differential expression analysis: t-test, Wilcoxon test, non parametric tests. Functional enrichment analysis with respect to Gene Ontology terms, KEGG pathways and specific a priori defined functional classes. 3.2 Gene function prediction. Supervised machine learning methods for the functional annotations of genes: Naive-Bayes classifiers, Neural Networks, Support Vector Machines. Ensemble methods for the integration of multiple sources of evidence. 4. Semi-supervised metohds. 4.1 Biomolecular networks analysis. Functional interaction networks and their modeling through graphs. Graph-based algorithms for the analysis of biomolecular networks: Guilt by Association, Random walks and label propagation algorithms. 4.2 Gene function prediction and disease gene prioritization as node label ranking problems in undirected graphs.