Understanding Disease Informatics System (UDIS): A Foundation to Integrate Exploratory Translational Data
By Martin Strahm, Head Data Science Basle,
F. Hoffmann-La Roche
Speaker for our 4th Annual Symposium:
Talk time: 10:00 am 27.3.2014
Talk abstract: "UDIS is the solution we built in-house to integrate –omics data in one place. The system brings clinical (human cohort) and pre-clinical (animal, cell lines) data from internal and public sources together. It serves as data source for simple visualizations suitable for many scientists and additionally via batch download for expert analysts.
The first challenge to build such a system is the huge amount of data that has to be stored and made accessible with reasonable performance. The second is the variety of data sources which store their data in different formats and use different vocabularies.
We could overcome the first challenge quite easily by bringing the right talents together. The second one is currently solved by a reasonably large investment in resources for data loading and curation, but would tremendously benefit from one industry wide standard."
Bio: "Martin received a master in High Energy Physics in Hamburg at the Deutsches Elektronen-Synchrotron (DESY). After this education in basic research he moved to Göttingen and received a Ph.D. in Biophysics at the Max-Planck-Institut für biophysikalische Chemie. He joined F. Hoffmann-La Roche Basle 1999 doing Bioinformatics. Since 2011 he has been the Local Area Head of the Disease and Translational Informatics group. During this time he has been leading a major project to develop a warehouse to integrate internally generated and external omics data types like sequence variations or RNASeq data. Since 2014 he is heading a Data Science group composed of 10 people with a variety of skills ranging from statistics, machine learning, cheminformatics, bioinformatics to information science."