EuResist

EuResist is an international project designed to improve the treatment of HIV patients by developing a computerized system that can recommend optimal treatment based on the patient’s clinical and genomic data.

The project is part of the Virtual Physiological Framework, funded by the European Commission . It began in 2006 with the formation of a consortium of several research institutes and hospitals in Europe and Israel . The consortium completed its commitment to the European Commission at the end of 2008, at which time the system became available online. A non-profit organization was consequently established by the main partners to maintain and improve the system.

In 2009, the EuResist project was named as a Computerworld honors program laureate. [1]

Background

AIDS is a disease caused by the HIV retrovirus , which progressively reduces the effectiveness of the immune system , leading to infections and ultimately death .

More than 30 different drugs exist for treating HIV patients. Antiretroviral drugs can disrupt the virus’s replication process causing its numbers to decrease dramatically. While the virus can not be eradicated completely, in small numbers it is harmless. A HAART , or HAART , is a highly active antiretroviral therapy . The main reason for this is that it is possible to prevent the development of mutations in the virus.

Thus an important consideration when choosing treatment for a patient is to prescribe which drugs to the particular patient’s viruses strands are most likely. One way to achieve that is to extract virus samples from the patient’s blood and test them against all possible drugs. Since this process is lengthy and costly, computerized systems have been developed to predict virus resistance based on its genotype. The treating physician samples virus genotype sequences from the patient’s blood and provides this data to a computerized system. The system then responds with drug recommendations.

Such systems are limited in accuracy, depending on the amount of data used for their creation, their quality and the richness of the mathematical models used for the actual prediction. Prior to EuResist, such systems have negatively impacted their accuracy: [2]

  • The amount of data used for creating the system was relatively small
  • This data was in vitro data: laboratory measures of the resistance of various strands of HIV to individual drugs. Such data are known to be inaccurate because they do not simulate the processes of a living organism.
  • They used a relatively simple mathematical prediction model

EuResist overview

EuResist Sought to create a more accurate prediction HIV treatment system by collecting a wide database of in vivo data (clinical and genomic records of real treatments of HIV patients and Their consequences), and by using an array of prediction models INSTEAD of just one.

The database was created by merging local databases of various clinics across Europe. This database is designed to be the largest of its kind in the world. For each patient, it includes various personal and demographic details such as gender, age, country of origin, genomic sequencing of HIV in the patient’s blood, records of the drugs prescribed, and the changes in the blood treatments.

This data Was used to process an array of prediction models, created by using various contemporary machine learning techniques, Among Them Bayesian networks , logistic regression , and others.

A web interface provides clinical and genomic data. This data is sent to the prediction engines, and the combined response, which is displayed to the physician, includes various provisions and a prediction of their effect on the amount of HIV in the blood.

The EuResist system was tested and compared with its predecessors. The developers of EuResist, who conducted this test, reported an improved performance over the previous state-of-the-art system.

History

EuResist started in 2006 as a consortium funded by the European Union as part of the Virtual Physiological Human FP-6 framework. The partners of this consortium were:

  • Informa Srl ( Italy )
  • University of Siena ( Italy )
  • Karolinska Institutet ( Sweden )
  • Universitätsklinikum Koeln ( Germany )
  • Max Planck Institute for Informatics ( Germany )
  • IBM Haifa Research Laboratory ( Israel )
  • MTA KFKI Reszecske- és Magfizikai Kutatóintézet ( Hungary )
  • Kingston University ( United Kingdom )

The Consortium completed its commitment to the European Union in late 2008, at which time the EuResist system became available on line. The prediction engines according to the principles of the prediction of the prediction of the prediction of the precipitate. As of mid-2010, an average of 600 queries are submitted to the EuResist system every quarter.

Recognition

On June 1, 2009, EuResist received a Computerworld honors program laureate award, a global program honoring individuals and organizations that use information technology to benefit society. [1]

References

  1. ^ Jump up to:b http://www.cwhonors.org/laureates/finalists2009.htm Computerworld honors program – laureates of 2009.
  2. Jump up^ M Rosen-Zvi et al. , Selecting anti-HIV therapies based on a variety of genomic and clinical factors, Bioinformatics 24 (13): i399-406, 2008http://bioinformatics.oxfordjournals.org/cgi/content/short/24/13/i399

Leave a Comment

Your email address will not be published. Required fields are marked *