Dr. Mattias Rantalainen from University of Oxford Visited WIPM

With the invitation of Professor Tang Huiru and Wang Yulan, Dr. Mattias Rantalainen from University of Oxford paid a 7-day academic visit to WIPM. During the time, he delivered excellent report on March 29 entitled “ Predictive Multivariate Latent Variable Models for Biomarker ”.
Mattias Rantalainen currently holds a Medical Research Council (MRC) Fellowship in Biomedical Informatics (2009-2012) and is based at the Department of Statistics at University of Oxford. Prior to his current position he was a postdoctoral researcher in Prof Chris Holmes' group at the Department of Statistics (Oxford) working on the data analysis work package for the MolPAGE consortium. He completed his PhD at Imperial College London where he developed novel multivariate pattern recognition methods with applications in metabonomics, working together with Professor Elaine Holmes and Professor Jeremy Nicholson. Mattias has an undergraduate degree in Engineering Biology (civ.ing. Teknisk Biologi) from Umeå University in Sweden.
In the report, Dr. Mattias Rantalainen pointed out what extent a non-linear model, kernel-based orthogonal projections to latent structures (K-OPLS), can provide enhanced predictive performance compared to the linear OPLS model. Their results reveal that the non-linear K-OPLS model provides improved prediction performance in three separate metabonomic data sets compared to the linear OPLS model. He also demonstrated how response-orthogonal K-OPLS components provide valuable biological interpretation of model and data.
At the end of the report, Dr. Mattias Rantalainen carried out productive communication and discussion with research staff and students and expressed the desire of deeper cooperation in future.