IBM Research       


IBM Research - Collaboration melds clinical and genomic data to cure diseases

IBM-Haifa works with Tel Aviv University on open source project to leverage anonymized medical data in two EU-based databases.

Michal Rosen_Zvi In one corner we have clinical data – a patient record of diagnoses and treatments collected over a period of time. In the other corner we have genomic data – a complete profile of the genetic material that defines the human species. How can researchers use both kinds of information to treat a range of illnesses, including cancer, HIV and cardiovascular disease?

Researchers in the machine learning group at IBM Research - Haifa and in the Edmond J. Safra Bioinformatics Program at Tel Aviv University are working on various projects that will weave clinical and genomic data together with the goal of curing disease, counseling patients to avoid health hazards and using bioinformatics databases efficiently.

The IBM and Tel Aviv researchers are working jointly under the auspices of IBM Open Collaborative Research (OCR), an open source project that supports software research between IBM and universities around the world. The OCR collaborators already have had access to clinical and genomic information housed in EuResist – an EU-based database of anonymized HIV patients along with their genotypes (cellular genetic makeup) and antiretroviral therapies – and in HyperGenes – an EU-based database containing clinical and genomic data from anonymized hypertension patients.

Avoiding problems with multi-use databases

The collaboration also examines problems inherent in using a single data set for a myriad of research purposes. As part of the OCR, the joint research team has developed a sophisticated method to correct for apparent inconsistencies and contradictions that stem from such multi-use activities. The so-called quality preserving database improves on earlier methods that overlooked details and subtleties when correcting for multiple testing by different parties. OCR researchers Hani Neuvirth-Telem, Udi Aharoni and Saharon Rosset have published their results in a paper about using public databases and in a paper for encouraging collaboration while controlling for false discovery.

The project researchers -- IBM OCR lead Michal Rosen-Zvi (pictured above), and Ron Shamir, Saharon Rosset and Eran Halperin from Tel-Aviv University -- organized a clinical genomic analysis workshop that hosted local academic and medical experts. The workshop held in June 2011 at the IBM Research lab in Haifa addressed issues in data mining, clinical-genomic data “fusion,” complex population structure and other related topics.

Open Collaborative Research is an IBM program that supports software research between IBM and universities. It aims to contribute to the growth of open source software. For more information, contact Steve Lavenberg, associate director, computer science, IBM Research.

Last updated on September 26, 2011.