IBM R&D Labs in Israel News

Diving into the gene pool

Joint IBM-Tel Aviv U research project to advance clinical genomic analysis

In an effort to advance clinical knowledge about the human genome, machine learning researchers from IBM are working on an Open Collaboration Research (OCR) project with scientists from Tel Aviv University in Israel. The unique collaboration will focus on learning more about the nature and treatment of disease based on clinical genomic analysis research.

The collaboration aims to advance knowledge about the structure of genomic data. The research will be led by researchers at IBM's Research Lab in Haifa and the Edmond J. Safra Bioinformatics Program at Tel Aviv University. The IBM Research OCR program funds joint projects with university researchers where collaborative innovation can benefit IBM and the general community.

Understanding genomic data

In recent years, Haifa researchers from the Machine Learning and Data Mining group have worked on connecting genomic data to clinical information, as part of the EuResist and HyperGenes EU research consortiums. These two projects examined how information technology can help improve and optimize the treatment of diseases such as AIDS and hypertension. Tel Aviv University researchers, led by Prof. Ron Shamir, are also leaders in this area, making collaboration between the two institutions mutually beneficial.

"Thorough understanding of the the structure/nature of genomic data is essential for the success of such studies," explained Michal Rozen-Zvi, manager of the Haifa Machine Learning and Data Mining group. Together with Shai Fine, the manager of the IBM Research - Haifa Analytics Department, she identified clinical genomic analysis research as a strategic domain for her team. That decision led to the current OCR project. At a kickoff meeting organized by IBM researcher Naama Parush-Shear-Yashuv, researchers from both teams presented their work.

Areas of research

The research collaboration with Tel Aviv U. scientists is examining a number of different areas.

  • IBM researcher Hani Neuvirth-Telem and Yaara Goldschmidt are working with TAU's Eran Halperin to discover a common genetic basis between various characteristics, or phenotypes, connected with a specific disease.
  • Udi Aharoni is working with Safra researcher (and former IBM Watson researcher) Saharon Rosset on a new approach for multiple hypotheses testing in connection with public databases, especially those containing genomic data.
  • IBM researchers Michal Ozery-Flato and Liat Ein-Dor are expanding previous work with Prof. Ron Shamir on cancer chromosomes and gene organization.
  • Project members in general are experimenting with the IBM team's bioinformatics data mining (BDM) tool, recently made available on alphaWorks, which uses machine learning and other techniques to improve information retrieval from life science data to see how it can be further improved.

"Several projects we have worked on in the past are similar to the work being conducted in the Safra Program. By bringing together more people to work together on these projects, we all benefit from the pooled talent, interest, and insight," explained Rosen-Zvi. She expects to see some concrete results at the end of the OCR project, which is funded for a one-year period with a possible extension.

The new research partnership with TAU is the third such engagement in Israel under IBM's OCR program. One of the other OCR efforts in Israel, a joint project with the Technion – Israel Institute of Technology and Rambam Hospital, is also in the healthcare domain.

The projects with TAU and the Technion are part of IBM's long-term healthcare research program, which in mid-2010 announced a plan to invest $100 million in scientific medical research over the next three years. Through these initiatives, IBM aims to make significant research advances and drive innovations that enhance patient care.


Joint IBM-TAU paper

The OCR project between IBM Research - Haifa and the TAU Safra Bioinformatics Program has already achieved some significant results. One of them is the joint paper by IBMers Ehud Aharoni and Hani Neuvirth and Safra researcher Saharon Rosset, "The Quality Preserving Database: A Computational Framework for Encouraging Collaboration, Enhancing Power and Controlling False Discovery." The article can be viewed on the IEEE Computer Society's Digital Library.