Feature - ViroLab: a grid-based decision support system for HIV infections

According to UNAIDS reports, 38.6 million people were living with HIV/AIDS in 2005, up from 36.2 million in 2003. Image courtesy of ViroLab

Medical doctors and researchers have more information at their fingertips than ever before. They can now study diseases from the DNA level all the way up to medical responses, thanks to the increasing availability of genetic information, extensive patient records, and large, high quality clinical and pharmacological databases.  

To coordinate this wealth of information medical doctors and researchers must have integrative technology that bridges existing gaps in multi-scale models, data fusion and cross-disciplinary collaboration.

It is here that distributed computing—and grid technology in particular—can play a crucial role in helping researchers to understand the processes involved in new biomedical e-science, from bioinformatics to heath informatics.

Virtualized tools; personalized treatment

ViroLab is part of the European Union Information Society Technologies effort to expand the reach of computing grid infrastructures into new scientific domains, aiming to create a collaborative virtual laboratory for grid-based decision support in viral disease treatment.

The prototype system targets improved understanding of HIV treatment in the increasingly common case of HIV drug resistance.

Using a grid-based service oriented architecture, ViroLab “vertically” integrates biomedical information relating to viruses (proteins and mutations), patients (e.g. viral load) and literature (drug resistance experiments), resulting in a rule-based decision support system for drug ranking.

This map shows the dynamics of an HIV population over five years of simulation. Yellow nodes show healthy individuals; red nodes are infected individuals, and black nodes are dead; the blue lines show links between individuals. These simulations give researchers an idea of what virus subtypes can be expected. Image courtesy of ViroLab

ViroLab system can be used to relate a patient’s viral genotype to a drug-susceptibility phenotype, using a distributed relational database containing a large number of phenotype-genotype pairs.

The software’s output is then further compared to data from clinical studies and to the relationship between the presence of genotype and the clinical outcome.

ViroLab also includes advanced tools for statistical analysis, visualization, modeling and simulation.

Initial ViroLab results indicate that the grid architecture prototyped for the use case of personalized HIV drug ranking is feasible, easily extensible, and highly applicable for end users.

Ease of use equals actual use

Data presentation and management is crucial in reducing barriers to adoption and actual usage by physicians, and here grid technology plays an important role.

ViroLab enhances its virtual organization by providing support for scientific collaboration, in order to process the variety of data and information generated from ViroLab applications, data providers and hospitals.

For instance, ViroLab supports the sharing of current and new drug rankings resulting from the new applications, collaborative validation of a new rankings, and feedback from experts via links to the ViroLab workflow engine.

The long term mission of ViroLab is to provide researchers and medical doctors with a virtual laboratory for furthering the study and treatment of infectious diseases. The ViroLab system remains under development and constant validation, with new resources being virtualized via web services and new functionalities being added from usability studies in a network of European hospitals.

Virolab involves twelve partners from eight European countries and makes use of the Enabling Grids for E-sciencE infrastructure.

Peter Sloot, University of Amsterdam

Peter Sloot is a professor of computational science at the University of Amsterdam as well as editor-in-chief of Elsevier Future Generation Computer Systems: The International Journal of Grid Computing: Theory, Methods and Applications.

Sloot will present new ViroLab results at the European Conference on Complex Systems, held 1-5 October at Dresdon in Germany.