Study of protein-protein interactions using primary structures

• Grant number: 11/11402-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: September 01, 2011
• End date: August 31, 2012
• Field of knowledge: Interdisciplinary Subjects
• Principal Investigator: Silvana Giuliatti
• Grantee: Ricardo Kazuhiro Kavanishi
• Host Institution: Faculdade de Medicina de Ribeirão Preto (FMRP). Universidade de São Paulo (USP). Ribeirão Preto , SP, Brazil

Abstract

The knowledge of protein-protein interactions is of great importance, because they are involved in most of the biological processes, being the basis of the cellular processes. However, most methods are limited by the need of previous knowledge of the protein in study, which is the case of some biochemistry techniques, like coimunoprecipitation, that needs informations of the biochemistry and functionality of the protein. This problem is intensified when the job needs us to study newly discovered proteins. Beyond the use of experimental techniques, many computational methods have been developed and applied, complementing the process of detection of the protein-protein interactions. Some of these methods also depends not only on informations about the sequence, but also of some previous knowledge as, for example, structural data and/or expression data. To solve this problem, algorithms to predict interaction between protein pairs, from their primary structure, have been developed. These algorithms make use of machine learning algorithms (ML), support vector machine (SVM). Therefore, the objective of this project is to study algorithms of Machine Learning applied to PPI, based on pairs of primary protein chains, in order to evaluate the viability of its application on cases of PPI, involving proteins that offers only the primary structure information, which is the case of newly discovered proteins. (AU)