[Seminar] Feature extraction and classification of biological sequences

Title: Feature extraction and classification of biological sequences

Speaker: Fabricio Martins Lopes

Affiliation: UTFPR, Brazil

Abstract: In recent years, because of the emergence of high-performance sequencers, there has been a major advance in the volume of DNA and RNA data. This has brought the “omics” sciences into the era of “BIG Data” or “e-Science”. Currently, a huge volume of data is available for genomic (DNA) and transcriptomic (RNA) analysis with diverse applications. Works of association and/or genomic selection, genetic transformation or genomic edition are using more and more “omics” information that needs innovative and precise strategies in research. However, the analysis tools of these “omics” data still need more attention, being necessary for the development of computational methods that can select, extract and classify the information for the knowledge discovery. It will be presented alignment-free method for feature extraction through complex networks to better synthesize and select the relevant information in biological sequences. It uses the principle of machine learning, in which the method, through an initial set of data, automatically learns to recognize patterns and uses this learning to improve the analysis and recognition of features that identify the class to which the sequence belongs.