Bridging chemistry and biology for pharma industry through kernel-based large-scale data fusion
Term: 7/2014 - 6/2016Project founded by JANSSEN RESEARCH & DEVELOPMENT, a division of JANSSEN PHARMACEUTICA N.V. Project supported by Belgium government. Research Partners:
Hasselt University, Center for Statistics, Belgium
Johannes Kepler University, Institute of Bioinformatics, Austria
General aim:In the different therapeutic areas of drug development, the growing dimensionality in chemical space and biological data (multi-target, multi-gene, multi-pathway) is highly challenging and will become an important bottleneck in the future decision making process. Therefore, it is critical to improve the fusion of disease mechanisms and disease phenotypes with the corresponding biological and medicinal chemistry decision cascades. The general aim of this project is to span the high-dimensional biological and compound space by ensuring that the different effect levels are fused (Disease <-> Gene set <-> Gene ranking <-> Protein target ranking <-> Compound ranking) with all available or specifically new experimental data. By solving the data fusion challenges of all those data types, we aim to improve the efficiency of phenotypic drug design (and diminish the explosion of follow-up costs). The final result will be a small number of potential biological or chemical hypotheses, that project teams can rely on and that are small enough to allow cost-efficient follow-up scenarios, either at the protein target identification or chemical probing level (hit enrichment).