Workpackage 08

Workpackage Objectives

• Establish predictors of differential response to noradrenergic vs. serotonergic antidepressants using integrative modelling of clinical and genomic variables. This will include candidate gene tests and systematic genome-wide exploration.
• Establish genomic signature of treatment resistant depression to provide a target group for the development and testing of novel treatments. This will also initially involve candidate genes and then a systematic genome-wide exploration. 
• Develop and test the sensitivity/specificity of prediction of response at individual level.
• Characterise the functionality of identified genetic variants associated with response to antidepressants at molecular and system level.

Workpackage Leads

Academic Lead: Prof. Peter McGuffin and Dr. Rudolf Uher, King’s College London, UK
EFPIA Lead: Dr. Jens Robert Wendland, F. Hoffmann - La Roche, Switzerland

Workpackage Partners

• H. Lundbeck A/S
• AstraZeneca AB
• GlaxoSmithKline Research and Development Ltd.
• Abbott Laboratories 

Synergy between academia and industry

The academy and EFPIA will complement each other in their skills and the type of samples available to them. The team at IoP-KCL has developed considerable expertise in the statistical analysis of pharmacogenetic data in the GENDEP project and of genome-wide data in large-scale case control studies. The GSK contributes expertise in gene set prioritising, based on a database developed by their bioinformatics group led by Dr. Michael R. Barnes. However, the most critical element in NEWMEDS is the pooling of datasets – no single industry or academic site has the ability to bring together data from 6,500 subjects including placebo-treated subjects. This will allow creating one of (if not the) largest combined datasets that will enable a powerful pharmacogenetic analysis, whilst controlling for non-specific factors using the placebo-treated group. Furthermore, the academy- and industry-lead datasets are complementary in many ways: e.g. the academy-led studies may be more inclusive and generalisable, but they lack placebo-control and thus are unable to distinguish predictors of treatment outcome from determinants of prognosis irrespective of treatment.