TEL AVIV, Israel—As it continues to pursue the promise of antibody-drug conjugates (ADC), Compugen Ltd. recently announced the discovery and selection of five candidate targets for ADC cancer therapy, which are the result of the company’s second focused in-silico discovery program. The targets are entering initial experimental validation, with preliminary results expected in the second half of this year.
Dr. Zurit Levine, vice president of research and discovery, says that most of the targets are expected to be effective against multiple types of cancer.
“In this discovery platform we looked for targets that have characteristics of an ideal target for ADC technology: distinct expression in cancers of various origins as well as in late-stage cancers, metastatic cancers and targets that are associated with poor clinical outcome, low expression in normal critical tissues and that are predicted to internalize,” she explains.
“These discoveries demonstrate significant advantages that our predictive discovery methodology offers compared with traditional discovery,” Dr. Anat Cohen-Dayag, president and CEO of Compugen, said in a press release. “First, since the required discovery platform for this program was built through the enhancement of our existing infrastructure, the total time to date has been less than a year from initiation of the program. Furthermore, this enhanced capability will now be available for future discovery programs. Second, since this second program utilizes a different basis for selection and prioritization of results than the first program, it further demonstrates the applicability of our core predictive infrastructure to multiple areas of high industry interest. Third, as demonstrated by both programs, each focused discovery program has the potential to yield multiple possible novel product candidates.”
Compugen’s ADC target discovery program, which began in 2013, used the same predictive discovery infrastructure as its earlier immune checkpoint program, as well as some algorithms and computational capabilities developed specifically for this program. In both cases, the objective was the identification of an appropriate set of proteins with the help of Compugen’s proprietary predictive human proteome, from which it could then select the proteins predicted to have the best chance of meeting the specific requirements of each program. Compugen began a collaboration with Bayer HealthCare in September 2013 to research, develop and commercialize antibody-based cancer immunotherapies against two novel immune checkpoint regulators discovered in its first in-silico target discovery program.
Levine says the success seen in Compugen’s first discovery program with immune checkpoint targets is “an exciting accomplishment for the company.”
“This accomplishment, achieved in a relatively short time, together with our ability to already demonstrate in well-accepted experimental systems that some of them have the required immuno-modulatory activity and the licensing of two of them to a large pharma company, clearly demonstrated the power of our predictive discovery capabilities,” she notes.
The advantages of these programs, Levine explains, are “the quality of the discoveries, the novelty of the resulting product candidates leading to the potentially first-in-class therapeutics and the relatively short discovery time.
“Our approach is predictive by its nature, and therefore enables the discovery of product candidates, some of which cannot be discovered through experimental discovery approaches. It took us over a decade to create our predictive and broadly applicable evolving discovery infrastructure, by utilizing the understanding of various key biological phenomena, such as alternative splicing, naturally occurring antisense, pseudogenes, gene fusions and more. We model these phenomena into the computer and integrate this knowledge together with information found in the public domain and proprietary data. The approach is unique and allows us to deal with biological complexity. Our various discovery platforms are integrated, which allows a competitive edge in comparison to single discovery approach or to high-throughput screening. The in-silico approach is based on computational models, which by definition improves with each discovery iteration, as experimental results are fed back into the models, improving them and enhancing the underlying algorithms.”
