Year of Graduation
Machine Learning Applications to Genome Annotation Problem
Big Data Systems
In this Master’s Thesis was developed and evaluated analytical programs that analyze genomic data in order to annotate unknown functional elements, in this case DNA secondary structures, based on the known experimental data from the Next Generation Sequencing experiments. The developed programs are based on the machine-learning algorithms, and in the frame of this Master’s Thesis we evaluated four: Random Forest, Support Vector Machines, Neural Networks, and Elastic Net. The algorithms were tested on big data of genomics publicly available from the International consortium project The Roadmap Epigenomics that collects data to study tissue differentiation. The developed programs can be used as modules in analytical systems used in the personalized medicine with the aim to find association of a patient’s genomic data with known genomic functional annotations.