Simulations Plus, Inc. (NASDAQ: SLP), a leading provider of simulation and modeling software for pharmaceutical discovery and development, announced that it has released Version 3.0 of its best-in-class ADMET Predictor software for predicting properties of molecules from only their structures.
Dr. Robert Fraczkiewicz, team leader for ADMET (Absorption, Distribution, Metabolism, Excretion and Toxicity) Cheminformatics for Simulations Plus, said: “This major upgrade to ADMET Predictor provides 14 new predicted properties as well as new retrained models for our previously top-ranked predictions. New predictions include eight models for metabolic inhibition and five for intrinsic clearance that were added to our previously released Enslein Metabolism Module. The initial Enslein Metabolism Module predicted rate of metabolism by certain enzymes, but not the potential for inhibition of metabolism that can cause significant drug-drug interactions with other drugs that are metabolized by the same enzymes. We’ve also added a new environmental toxicity prediction based on data originally generated by the U.S. Environmental Protection Agency.”
Dr. Fraczkiewicz continued: “In ADMET Predictor 3.0, we’ve incorporated the innovative technology we developed under last year’s SBIR grant and subsequent internal R&D to rapidly calculate partial atomic charges. ADMET Predictor’s new partial atomic charge calculations provide over 50 new descriptors that serve two purposes. One is to train mathematical models that correlate predicted properties with descriptors. The second is to provide a new visualization tool that lets chemists see charge distributions of various kinds directly on a drawing of the molecule – an important tool that gives chemists insight into why molecules behave the way they do and how to modify their structure to achieve certain goals. In the past, calculating these charge descriptors would have required as much as a day or more per molecule. Our breakthrough methodology calculates these charges at the rate of hundreds of thousands of molecules per hour, with a similar level of accuracy.”
“In addition, we improved ADMET Modeler to better select descriptors during model building,” added Dr. Fraczkiewicz. “This resulted in new models with significantly fewer descriptors, an important development because models that cover the widest range of new chemistry are generally those with the smallest number of descriptors. Another enhancement to ADMET Predictor is the addition of a state-of-the-art graphics engine called Miner3D™, which allows the scientist to visualize multidimensional relationships among properties and descriptors quickly and conveniently.”
Walt Woltosz, chairman and chief executive officer of Simulations Plus, said: “In silico(computer) property predictions provide chemists with the fastest possible means of filtering out poor molecules from large molecular libraries as they search for new medicines, and now ADMET Predictor’s top-ranked predictions are better and cover even more properties than previous versions. The tight integration we now have between ADMET Predictor and our ClassPharmer™ software product means that the de novo design capabilities of ClassPharmer will now also be enhanced, allowing chemists to automate the generation and screening of new molecular structures to a new level. The potential number of new drug-like molecules is said to be more than the number of grains of sand in the world, yet most of them will never become drugs because of undesirable properties. ADMET Predictor and ClassPharmer are at the forefront of their technologies for increasing the productivity of pharmaceutical scientists in this exceedingly difficult task.”