Multitask matrix completion for learning protein interactions across diseases
Four our next meeting on 3/28/2016 we have selected Multitask matrix completion for learning protein interactions across diseases by Kshirsagar et al. The abstract is as follows.
Disease causing pathogens such as viruses, introduce their proteins into the host cells where they interact with the host’s proteins enabling the virus to replicate inside the host. These interactions be- tween pathogen and host proteins are key to understanding infectious diseases. Often multiple diseases involve phylogenetically related or bio- logically similar pathogens. Here we present a multitask learning method to jointly model interactions between human proteins and three different, but related viruses: Hepatitis C, Ebola virus and Influenza A. Our multi- task matrix completion based model uses a shared low-rank structure in addition to a task-specific sparse structure to incorporate the various in- teractions. We obtain upto a 39% improvement in predictive performance over prior state-of-the-art models. We show how our model’s parame- ters can be interpreted to reveal both general and specific interaction- relevant characteristics of the viruses. Our code and data is available at: http://www.cs.cmu.edu/~mkshirsa/bsl_mtl.tgz
We look forward to seeing all who can come. Feel free to begin our discussion in the comments section below.