| Date |
Sep 25, 2018 |
| Speaker |
Jiangning Song,
Senior Research Fellow and Group Leader, Monash Biomedicine Discovery Institute (BDI), Monash University, Melbourne, Australia
|
| Title |
Harnessing the power of machine-learning techniques to address sequence classification problems in the era of biomedical big data
|
| Abstract |
Recent advances in high-throughput sequencing have significantly contributed to an ever-increasing gap between the number of gene products (‘proteins’) whose function is well characterised and those for which there is no functional annotation at all. Experimental techniques to determine the protein function are often expensive and time-consuming. Recently, machine-learning (ML) techniques based on statistical learning have provided efficient solutions to challenging problems of sequence classification or functional annotation that were previously considered difficult to address. In this talk, by combining our recent research progress, I will highlight some important developments in the prediction of two representative sequence labeling problems in computational biology, i.e. i.e. ‘target substrate labeling’ and ‘active site labeling’, based on the high-dimensional, noisy and redundant information derived from sequences and the 3D structure. I will illustrate how ML methods can extract the predictive power from a variety of features that are derived from different aspects of the data can contribute to the model performance.
|
|
アブストラクト