Learning from heterogeneous data sources: an application in spatial proteomics

We describe an application of transfer learning, where we combine experimental mass spectrometry-based spatial proteomics and freely available data from repositories to reliably infer sub-cellular localisation of proteins. More details - http://biorxiv.org/content/early/2015/07/07/022152; software - http://bioconductor.org/packages/release/bioc/html/pRoloc.html.