Fast feature selection method for large-scale analysis of local sequence/structure/function relationships in proteins
Category
- Completed
- DICP Tool Prototype Trial for Integrated Database Analysis
- Projects funded in FY 2013
Name and affiliation of Research Director
EBINA Teppei
Researcher, Brain Science Institute, RIKEN
Outline of R&D
Detecting amino acid sequence features of protein local structures and functional sites allows us to discover the relationships between protein sequence/structure/function. Interest in the fast search of the features has increased recently with rapid growth of protein databases as feature selection from very large data sets requires enormous computational costs. Here, we will develop a fast method of the feature selection by clustering amino acid sequence fragments using recently developed BOOL (Binary cOding Oriented cLustering) algorithm and will examine whether our method can identify "optimal" sequence features of protein local structures with faster computational time than other methods.
Main database(s) subject to research and development
- RVSB
Period of research and development
September 2013 to January 2014
