Machine learning for protein function prediction
Karin Verspoor (NICTA)
NICTA SML SEMINARDATE: 2012-04-19
TIME: 11:15:00 - 12:00:00
LOCATION: NICTA - 7 London Circuit
CONTACT: JavaScript must be enabled to display this email address.
ABSTRACT:
This talk will describe some recent applications of machine learning that we have exploring for the purpose of predicting protein function. We have developed an approach to protein function prediction that incorporates features derived from the biomedical literature with other sequence or protein-level information from databases in a structured SVM framework. In addition to describing this specific application, I will also introduce the context for text mining of the biomedical literature more broadly. If there is remaining time, I will finally describe some methods for learning of patterns for information extraction of specific protein-related biological events from the biomedical literature that I have been exploring with my collaborators.
