MORE ABOUTAlfonso Valencia Herrera
Martin Krallinger, Miguel Vázquez, Andrés Cañada, José María Fernández and Andrea Nicole Doelker, Centro Nacional de Investigaciones Oncológicas.
Precision Medicine is considered the cornerstone of future medicine. In particular, the combination of genomic information and patient information is the current paradigm in biomedical research.
Needless to say that the development of Precision Medicine depends critically of the availability of the adequate computational systems able to handle very large quantities of complex heterogeneous information (Big Data) and to reason on them.In this project, we propose to explore the application of Cognitive Computing, to Precision Medicine, with particular emphasis in the application of Deep Learning strategies. This is an essential step given the characteristic lack of large collections of labelled data specific of this area.
The development of Unsupervised Learning strategies is considered a key research topic for the development of Cognitive Computing, while Precision Medicine is seen as a particularly adequate environment for its application.To make this ambitious project possible we will take advantage of two of our systems that already in use in the context of the CNIO Precision Medicine initiative. The Rbbt framework provides the cancer genome analysis capacity and the Melanomamine system is used to extract biomedical information on Melanoma from scientific text.
These systems will provide organized and processed input data that will tremendously facilitate (and to large degree make possible in time and form) the training of the Cognitive Computing system. Furthermore, these two systems will provide a baseline for the assessment of the results of the Cognitive Computing systems. This project will be developed in the framework of the collaboration between CNIO and IBM, which makes possible the use of the IBM Watson system.