The technology will be presented at the annual conference of the life sciences industry in Israel - ILSI-Biomed 2010. The findings were published in the scientific newspaper Journal of Chemical Information and Modeling -
Drug discovery and development is a high-risk process, in which only one molecule out of 10,000 candidates gets to reach the market. When you take into account all the potential drugs abandoned along the way, the cost of developing one successful drug can reach a billion dollars, and the development process can take 15-20 years. For this reason, one of the most prominent trends in the pharmaceutical industry today is the integration of computational methods for predicting and locating promising molecules already in the early stages of drug development. A development process that focuses on quality potential candidates is expected to reduce the dropout rate and increase the degree of success significantly.
Applision, the research development company of the Hebrew University, unveils a new computational platform for drug discovery. The method will be presented by the inventor of the technology, Professor Amiram Goldblum from the Drug Research Institute of the Hebrew University of Jerusalem, on Tuesday, June 15, in the session discussing technology commercialization at the ILSI-Biomed 2010 conference, which will be held at the David Intercontinental Hotel in Tel Aviv. The technology is based on a unique algorithm for identifying molecules with the highest probability of having a desired biological activity. The platform has an extraordinary ability to predict the most promising molecule as well as to find a set of successful alternative molecules, thereby creating a sort of targeted library of potential drugs with desired properties. The platform is also able to rank the molecules according to their ability to react with certain proteins in the body or according to their expected effect on various diseases.
The new platform provides an enrichment of up to 5,000 times for molecules with biological activity, and it has already been tested in laboratory experiments. In one experiment, a database containing 2.5 million small molecules was fed into the platform, and the technology looked for potential inhibitors of a certain enzyme that could be used as drugs for Alzheimer's disease. Out of the 2.5 million starting molecules, the platform was able to locate 10 molecules that were supposed to have the required biological properties. Of these, 5 candidates were tested in the laboratory, and 3 indeed had the desired biological activity. The three molecules were not known before and were not patented. In another experiment,
The technology helped design a protein drug to treat CML leukemia. From a pool of 1080 protein sequences, a total of 10 were selected for laboratory experiments, and 6 of them did inhibit the growth of the cancer cells in culture. The findings were published in March this year in the scientific journal Journal of Chemical Information and Modeling
"The powerful predictive ability of the new computational platform will help to shorten the pre-clinical phase in drug development from 3-4 years to 2-3 months," said Yaakov Michlin, CEO of Amision. "In fact, a 10% improvement in the ability to predict the failure of a potential drug could save millions of dollars in the development cost of each drug.
One response
I think this is an exciting achievement.