Active learning in hierarchical classification of transposable elements
| Grant number: | 16/12489-2 |
| Support type: | Scholarships in Brazil - Master |
| Effective date (Start): | January 01, 2017 |
| Effective date (End): | August 31, 2018 |
| Field of knowledge: | Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques |
| Cooperation agreement: | Coordination of Improvement of Higher Education Personnel (CAPES) |
| Principal Investigator: | Ricardo Cerri |
| Grantee: | Felipe Kenji Nakano |
| Home Institution: | Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil |
| Associated scholarship(s): | 17/19264-9 - Active learning in hierarchical classification of transposable elements, BE.EP.MS |
Abstract Transposable Elements (TEs) are DNA sequences that can change its location within a cell's genome. They contribute directly to the genetic variety of species. Besides, their transposition mechanisms can affect the functionality of genes. The correct identification and classification of TEs play a central role in comprehension of genomes. Generally, identification and classification of TEs are performed using tools that utilizes homology, by comparing a sequence to many sequences from a labeled TE database. This method is limited, since the homology ignore sequences' biochemical properties and relations among different TE classes. Since the literature proposes hierarchical taxonomies to classify TEs according to classes and subclasses, this project aims to develop new classification methods employing Machine Learning, considering hierarchical relationships among different classes. More specifically, artificial neural networks trained using Deep Learning concepts will be investigated. As the first step, datasets will be constructed from TEs sequences already identified. In order to build such datasets, Bioinformatic tools, capable of identifying the presence of signatures and biochemical characteristics, will be used. Also, different strategies will be used to convert sequences to attributes suited for Machine Learning. Afterwards, the datasets will be structured in a hierarchical fashion, according to TEs families and superfamilies. The new proposed classification methods will be compared to state-of-art methods from literature, and evaluated using measures specifically designed for hierarchical classification problems. (AU) | |