Background


Introduction

Over the last few years, researchers have realized that many dimensions of biology and medicine aim to understand and model the informational mechanisms that support more precise clinical diagnostic, prognostic and therapeutic procedures. As long as data grows exponentially, novel BMI approaches and tools are needed to manage the data. Although researchers are typically able to manage this information within specific, usually narrow contexts of clinical investigation, novel approaches for both training and clinical usage must be developed. Most clinicians are not familiar with this kind of information; how to use it can be controversial, so informatics tools must be developed to help in this process. If not, it is likely that the inclusion of genomic information in clinical practice will be delayed for years given the reluctance of physicians to use complex, often incompletely tested, non-standardized, and very partially connected biological information.

After the Human Genome was completed and other "omics" technologies began to proliferate and have an impact, biomedical scientists and practitioners proposed to use the new wealth of biological information, such as SNPs, microarray data, biomarkers, with their ever-increasing data sets into clinical practice. Researchers have expressed the hope that new findings could have immediate clinical application, leading to new visions of medicine, such as "genomic medicine" or "personalized medicine". In such new scenarios, new computational infrastructures, based on Grid technologies, will be needed.

The vision of a large, Grid-based, distributed computing environment, where researchers share the processing capabilities of many computers is particularly relevant in the biomedical area. Over the last decade, demanding tasks such as microarray analysis, mining very large biomedical databases, data integration, simulation, image processing and others have raised an increasing interest in the biomedical community. Such interest also led to the creation of new infrastructures and organizations in the biomedical area, based on Grid technologies.

Many initiatives have been launched over the last few years to build such Grid infrastructures both in the Europe and USA. In Europe, the most relevant organization in the area, called HEALTHGRID - member of this consortium-, has led this dissemination effort. It regularly organizes conferences and has supported the writing of a White Paper in the area, elaborated by a group of experts at a European level. HEALTHGRID, with the active participation of the European Commission, has been promoting the development of new Grid initiatives in Europe. In the USA, an initiative called the BIRN, Biomedical Informatics Research Network - although BIRN it is not purely a BMI initiative, at least as the term BMI has been considered in Europe-, has been launched by the National Institutes of Health to address the feasibility of applying Grid technologies in biomedicine. As regards to Europe, there are many examples of GRID-based projects, such as GenoGrid, Gripps, MediGrid, GLOP, MyGrid, eDiamond, and others, funded by the European Commission Based on these achievements, members of ACTION-Grid have introduced in a recent survey paper, published in the June issue of the IEEE Engineering in Medicine and Biology Magazine [Maojo and Tsiknakis FP7-ICT-2007-2 Annex I ACTION-Grid 14/05/08 Page 7 of 93, 2007]1 some modifications, proposing the areas and demanding computational issues where

Grid application to genomic and personalized medicine

Grid technologies can support genomic and personalized medicine, such as, for instance:
  1. "Omics" areas. For instance, for identifying genes and proteins or the automatic annotation of genomic information.
  2. Building virtual models of molecules, cells and organs. Building a "virtual human" project is a natural evolution of past efforts such as the US NLM Visible Human Project, incorporating new knowledge on systems biology, physiology and the "omics" areas and modeling techniques.
  3. Molecular imaging, developing "in vivo" visualisations of cellular and genetic processes. Molecular imaging developments pursue quantitative and non-invasive studies of diseases at the molecular level. Grid can provide the processing power needed by this area.
  4. Image processing and management. Applications such as 3-D modeling and visualization, prediction of 3-D protein structures or image storage and transmission require an underlying computing infrastructure that traditional networks and techniques cannot support.
  5. Pharmacogenomics. The design of new drugs requires advanced computational methods and tools. New models of clinical trials will require in-silico simulation, reducing the time needed in classical approaches.
  6. Surgical planning and simulation. Multimodal image fusion and real-time visualization and image manipulation are needed for creating realistic models of surgical interventions, for use in training, diagnosis and surgery. In this scenario, Grid computing might contribute to facilitate research.