Title: Deep Learning for Biological Big Data Analysis
Abstract: This talk will use protein folding as an example to demonstrate the application of deep learning to biological sciences.
Ab initio protein folding is one of the most challenging problems in Computational Biology.
It has been studied for several decades, but not much progress has been made until recent years when a large number of protein sequences and a reasonable number of protein structures are available.
To make full use of available sequence and structure data, we have developed a novel deep learning (DL) method to learn to fold proteins from existing sequence and structure databases.
Our DL method is composed of two deep residual neural networks (ResNet).
The first ResNet conducts convolutional transformation of 1-dimensional protein features to capture sequential context of one residue
and the second conducts convolutional transformation of 2-dimensional features to exploit higher-order residue correlation and global information.
Experimental results suggest that our DL method doubles the accuracy of pure co-evolution methods and can fold many more proteins than ever before while running on a single workstation.
By contrast, previous methods have low folding accuracy even if running on hundreds of CPUs.
Our DL method also works well on membrane proteins, a specific type of proteins that are very important for drug discovery.
简介：Dr. Jinbo Xu is a full professor at the Toyota Technological Institute at Chicago, a computer science research and educational institute located at the University of Chicago. Dr. Xu’s research lies in machine learning, optimization and computational biology. He has developed a very popular software package RaptorX (http://raptorx.uchicago.edu) for protein structure prediction. Dr. Xu is the recipient of Alfred P. Sloan Research Fellowship, NSF CAREER award and 2018 PLoS Computational Biology Research Prize.