Future CDD project plans and project applications

Also in the near future CDD will remain focused on structural bioinformatics (protein flexibility in drug design, data-driven pharmacophore modeling, and structure-based molecular design) and integrating bioinformatics and cheminformatics technologies to support omics and translational medicine research. Title of recently (approved) CDD research proposals are

  • A Systems Bioinformatics Approach For Evaluating And Translating Drug-Target Effects In Disease Related Pathways
  • Translational Medicine Informatics: Application of adverse drug effect data in discovery research
  • It's good to be flexible - taking protein flexibility in SBDD to the next level (VENI project, Dr. Nabuurs)

 

The project proposals are follow up studies of the ongoing CDD projects described in section 5.1, 5.2, 5.3, 5.4 and 5.5 and will form the core of future research in CDD.

 

CDD, Systems Bioinformatics for Drug Discovery

A lot of newly developed drugs fail in clinical trials because of lack of efficacy for the anticipated indication or unexpected toxicity (Kola, 2004). Apparently, it remains hard to establish a clear link between antagonism or agonism of a specific target and its effect in human disease and its target related toxicity.

One important reason for these high attrition rates is the often underestimated complexity of protein function in higher order biological systems, in which numerous protein-protein interactions, feedback loops and redundancies play a role. The collection of canonical pathways that can be found in public databases and commercial tools do not adequately address these issues because they are mainly a reflection of experimental data that are obtained from isolated cell lines and tissues. They address mostly, the signaling events that lead to binding of transcription factors to the DNA, but do not detail the pleiotrophic effects that arise downstream from the induced transcriptional program, which are most important in eliciting the systems response to the signaling events and may determine in large part, the efficacy and toxicity of a drug.

Another problem is that the target validation and drug efficacy studies are mostly done in animal models, and that the translation of these animal datasets to a human setting is extremely difficult to be realized. Most comparative genomics tools are aimed at studying conservation of single genes or gene families, whereas tools that computationally address orthologous biology, i.e. conservation of the entire pathway or pathways in which the target is involved, are scarce. This seriously impedes the output and effectiveness of translational research from pre-clinical to clinical studies.

We want to develop a systems biology approach that addresses the above problems and apply that to rheumatoid arthritis (RA), an inflammatory autoimmune disease that affects a large part of the population worldwide and for which there is still a high unmet medical need.

 

Scientific Approach: Create biological networks related to rheumatoid arthritis:

Sets of genes that are related to RA will be collected from microarray data from individuals with various forms of RA and that have received various drug treatments. These data are partly available at Schering-Plough, and will be complemented with data from public repositories. From these genes, networks will be built based on, amongst others, the patterns of co-regulation, shared regulatory elements in the promoter region, concordance with protein-protein interaction data and co-occurrence in the primary literature (Frijters, 2007). This systems biology approach will yield networks that represent the biology that is more focused on the pathophysiology of RA than is captured in canonical inflammatory pathways.

 

Assess orthologous biology:

The networks that are generated for the human biology will be compared with networks based on data from animal models that are believed to be a representatation of the human disease of interest, such as the mouse collagen induced arthritis and the mouse acute inflammation model. Furthermore, we will develop methods that estimate conservation of entire pathways based on the number of orthologs in the pathways, the overall level of sequence conservation and chromosome synteny (Hulsen, 2006). Densely connected genes with orthologs in multiple organisms represent important genes and may be attractive targets for therapeutic intervention.

 

Integrate and visualize the data:

Methods will be made available and/or developed, to visualize the gene networks, the underlying literature links together with the results from the conservation analysis and relevant gene annotation in SVG format and via links to other networks analysis software such as Cytoscape, including interactive webservers, in which users can create and analyze networks based on their own microarray data..

 

The combined approach of generating biological networks using a systems biology approach together with integration of conservation analysis and important gene annotation will allow for quicker and better experiments aimed at evaluating multiple targets and drugs for further clinical development. This will be a first step to reduce the high attrition rates associated with drug development.

 

References:

2007 R. Frijters, S. Verhoeven, W. Alkema, R. van Schaik, and J. Polman. Literature based compound profiling: application to toxicogenomics. Pharmacogenomics, 8, 1521-1534.

2006 T. Hulsen, J. de Vlieg, P.M. Groenen PhyloPat: phylogenetic pattern analysis of eukaryotic genes. BMC Bioinformatics, 7: 398.

2004 I. Kola, J. Landis Can the pharmaceutical industry reduce attrition rates? Nature Reviews Drug Discovery, 3: 711-715.