Machine Learning Algorithms for Protein Structure Prediction

Mr. Jianlin Cheng
Thursday, December 8, 2005
1:30PM - CSB-232

Abstract


The amino acid sequence of a protein specifies its three dimensional (3D) structure, which determines the protein's function. With the exponential growth of protein sequences without determined 3D structures in the post-genomic era, the prediction of protein structure from protein sequence has become one of the most fundamental problems in structural and functional bioinformatics.

Machine learning methods provide powerful means for pattern recognition and have been widely used in computational biology and bioinformatics. In this talk, I will present several machine learning algorithms (e.g., neural networks and support vector machines) for the prediction of protein structure and structural features from primary sequences. In particular, I will focus on the three-stage prediction of protein beta-sheets using neural networks, alignments, and graph algorithms and a novel machine learning information retrieval framework for protein fold recognition - the key step for predicting protein 3D structure.

Short Bio


Jianlin Cheng is a PhD candidate (advisor Pierre Baldi) in bioinformatics in the Donald Bren School of Information and Computer Sciences at the University of California, Irvine. His research interests include algorithms and applications for bioinformatics, systems biology, machine/statistical learning, and data mining. He has an MS in computer science from Utah State University and a BS in computer science from Huazhong University of Science and Technology in China.