Molecular modelling in cancer research

Advances in computing are vital to cancer research. They open up opportunities to discover new drugs from the molecular basis of disease.

Speeding up virtual molecular modelling in cancer research

Advances in computing are opening up new opportunities in cancer research, from virtual screening through to new resources to attract postgraduate researchers, and increasing the number and range of inter-disciplinary projects running simultaneously.

For researcher Dr Jack Flanagan and his team at The University of Auckland’s Cancer Society Research Centre (ACSRC), a lack of computing resources became a major obstacle holding back progress.

“We had one computer to do everything for everyone. It was under-powered, crashed constantly and it could take weeks to run the library,” says Dr Flanagan.

Advances in computing are vital to cancer research. They’re opening up opportunities to discover new drugs from the molecular basis of disease by a technique called virtual screening. This is proving to be a cost-effective method of identifying compounds that block the action of many proteins suggested by biology as important factors in a wide variety of diseases.

Dr Flanagan and his team use 3D computer modelling to show how drugs might interact with their target protein molecules. Virtual screening scales up this process to look at how hundreds of thousands of compounds might bind to the protein and then ranking the results based on some assessment of complimentarity.

Initially, a virtual screening campaign against one target protein with a library of approximately 100,000 compounds would take several weeks, but did represent the first steps in developing a local discovery platform that could offset the lack of a high throughput chemical screening facility in New Zealand.

The team needed to move on and with funds from The University of Auckland’s Biopharma Theme Research Initiative they turned to the Centre for eResearch for help.

A solution from the Center for eResearch

eResearch co-Director Nick Jones worked with Dr Flanagan to design a solution based around computing power and BeSTGRID to support computational chemistry research.

“Working with Jack, we’ve been able to quickly pattern his research processes, and provide key computational and data systems tailored to meet his needs,” says eResearch co-Director Nick Jones. “This lets Jack concentrate on his research and collaborations, providing an enhanced and more efficient research environment for drug discovery.”

The result:

  • Scaled-up computing resources to support new virtual screening campaigns
  • A streamlined computational workflow supported by an easy-to-use application
  • Increased throughput capacity that allows regular screening of 100,000 compound libraries within 5 to 7 days.
  • Ability to enable other researchers at The University of Auckland with resources not available anywhere else

Benefits to Dr Flanagan and his team

The most immediate benefit has been in increased computing capacity, with the following key enhancements to research workflow:

  • More computing scale and power, by accessing BeSTGRID’s faster CPUs
  • Multiple researchers can access this resource at the same time
  • Access to secure storage space that allows sharing of result sets
  • Reduced search times that are improved with BeSTGRID hardware upgrades
  • Increased capability through BeSTGRID
  • Use of more complex software that is computer resource intensive

Access to BeSTGRID and collaboration with the Centre for eResearch directly increases ACSRC’s research capability.

“We can perform larger virtual screening analyses more routinely, and the results of these are filtering into some of the projects within the centre by providing another source of ideas for the medicinal chemists doing the drug development”, says Dr Flanagan. The extra computing resource has increased throughput and also enables the team to start using more detailed calculations that attempt to better model the interaction between drugs and their target proteins.

Enhancements offered by the virtual screening service are creating opportunities for more collaboration, and as research data increases exponentially in volume, new approaches are required for sharing, accessing and analysing data. Without specialist support, researchers are often unable to overcome these challenges. Recognising this challenge, the screening service supports collaboration through use of the secure and highly scalable BeSTGRID Data Fabric.

This has all been achieved with easier access via a researcher’s desktop without their needing to learn complex command lines and code-like interfaces. This application is adaptable for expert and non-expert users so provides a pathway for researchers to explore greater functionality if they so desire.

 “What all this means is that we can expand the types of questions we ask, such as how do different high throughput screening technologies compare, do computer based methods reflect the outputs of experimental methods, where are the strengths and weaknesses of the methods, these are possible to address if we can screen tens to hundreds of thousands of compounds routinely and it has scalability,” says Dr Flanagan.

Scalability

Now the prototype is running, Dr Flanagan is working to establish a repeatable processing pipeline which can be used across labs and institutions. “With further upscaling of the computational platform, we could process samples of 50,000- 100,000 candidates within one day, offering huge benefits in terms of research efficiency.”

Two ACSRC projects are supported by the current BeSTGRID virtual screening platform, and ‘active’ compounds have been discovered. With the development of this platform the number of projects using it has tripled across the University, and includes engagement with other researchers within the ACSRC, as well as the Maurice Wilkins Centre for Biodiscovery, the University of Auckland Faculty of Medicine and Health Sciences, and the Faculty of Science. Research initiatives between the ACSRC and international collaborators including Cancer Research Therapeutics (United Kingdom) and the Garvan Institute (Australia) also benefit from this platform by its contribution to lead development programmes.

Where to from here?

Nick Jones believes this infrastructure shows the transformative power of eResearch and represents a new capability in New Zealand science, particularly where there is potential for exploitation of intellectual property.

“Many other researchers and groups across New Zealand could benefit greatly from a similar modest investment to leverage large increases in research productivity and intellectual property generation,he says.

Over the summer of 2010 – 2011 extension of the service has been prototyped within a Summer of eResearch student scholarship programme, which has paired the virtual screening team with a team of three summer scholarship students to investigate a new problem: how to analyse the increasingly large result sets coming out of the virtual screening service. This is leading to new collaborations with computer scientists and bioinformaticians and statisticians.

“This nexus of innovating between science groups is where eResearch has its most valuable role to play,” suggests eResearch co-Director Nick Jones. “Many opportunities exist to overcome the challenges of working across disparate research disciplines, to act as a catalyst for new approaches, and to help researchers deal with increasing complexity in their daily work life. “

The Centre for eResearch

The Centre designs cyber infrastructures to supporting effective digital research communities through hardware/software, computational semantics, information science methodologies, and advanced research networks.

BeSTGRID

eResearch connects researchers to BeSTGRID (Broadband-enabled Science and Technology GRID) a data fabric coordinating leading universities and organisations enabling collaborative research on faster time lines and with greater computational power and data intensity. BeSTGRID has the capability to build a national eResearch infrastructure to support different research communities.

KAREN (Kiwi advanced research and education network)

KAREN is a super fast (10Gb/second) private global internet of 200 million users around the world that enables researchers to add a third pillar to theory and observation – one of calculation, simulation and modelling. New research is data intensive across creation, storage, analysis and visualisation, and KAREN enables this data to be shared, accessed and mined reliably and efficiently.

ACSRC

The ACSRC is a research centre in the School of Medical Sciences, jointly managed by the University and the Auckland Cancer Society.  Its major focus is drug development, which is often carried out in collaboration with both University and commercial partners.

For more information:

Nick Jones
Co-Director eResearch
Centre for eResearch
Ph +64 9 923 9748
n.jones@auckland.ac.nz
www.eresearch.auckland.ac.nz