Applications are invited for a fellowship in machine learning for personalized medicine, to be funded by the Marie Curie Initial Training Network MLPM2012 of the 7th Framework Programme of the European Commission. MLPM (http://mlpm.eu) is a consortium of several universities, research institutions and companies located in Belgium, France, Germany, the Netherlands, Spain, Switzerland, UK, and in the USA. MLPM offers an excellent training environment in the research field at the intersection of machine learning and medicine. It includes several academic labs with expertise in statistical genetics or machine learning, and private companies that are active in this field. The recruited fellow will be based at award-winning start-up company Pharmatics in Edinburgh, UK. The fellow will visit other nodes and attend training events of the network, in particular the annual summer schools on Machine Learning for Personalized Medicine. The immediately available position is fully funded (100% employment) until 31/12/2016 according to the Marie Curie programme, which offers a highly competitive and attractive salary.
Research project: The recruited fellow will develop novel machine learning / bioinformatics methods for predictions of complex disease-related phenotypes, disease subtyping, and patient stratification based on heterogeneous high-dimensional molecular measurements and clinical/environmental variables. The methods may be extended to longitudinal (time series) phenotypes. Effects of possible shifts between training and test distributions will be considered. The methods will be investigated for several clinical datasets and build on the current understanding of targeted clinical indications. If the new methods outperform current clinical models for stratification and prediction of diseases and related outcomes, they may be investigated further in prospective clinical studies and eventually packaged as medical devices.