Simulations Plus Announces Release of ADMET Modeler™ 1.0 and Upgrade of ADMET Predictor™

Software: ADMET Predictor®
Division: Simulations Plus

Simulations Plus, Inc. (AMEX: SLP), the leading provider of ADMET absorption simulation and neural net structure-to-property prediction software for pharmaceutical discovery and development, announced today that it has released version 1.0 of ADMET Modeler™, formerly known as QMPRchitect™, and an upgraded version of its ADMET Predictor™ software that was released last December.

Ron Creeley, vice president of marketing and sales of Simulations Plus, said: “As announced earlier, we have changed the name from QMPRchitect to ADMET Modeler to better communicate the purpose of this software package to customers and investors. In addition, ADMET Modeler is a companion program to ADMET Predictor, so the two names are now more easily identified with their purposes. ‘ADMET’ or ‘Absorption, Distribution, Metabolism, Excretion and Toxicity’ is a common acronym in the pharmaceutical industry. ADMET Modeler builds predictive mathematical models for ADMET properties of molecules from their structure. In fact, all of the artificial neural network ensemble models in ADMET Predictor were built with ADMET Modeler. The improvements to ADMET Predictor include new and improved solubility models as well as faster execution and intuitive graphics for our world-class pKapredictor.”

Boyd Steere, product manager for ADMET Modeler, added, “With the release of ADMET Modeler 1.0, we’ve added features that our users have requested, and we have extended the model-building capabilities to include not only artificial neural network ensembles, but also support vector machine ensembles. Support vector machines are one of the newest mathematical methods for building correlation models from data, and this capability was requested by a number of customers over the past two years.”

Robert Fraczkiewicz, product manager for ADMET Predictor, said: “The upgrade for ADMET Predictor, version 1.2.0, adds two new solubility models with greatly enhanced scope that provide a comparison of predicted solubility from models built on large datasets that include both drug-like molecules and other chemicals, and solubility predicted from models built only on actual drug molecules. Better graphics now provide an excellent depiction of microscopic ionization states that result when molecules ionize. Ionization is a crucial consideration with drug molecules because it can cause large changes in other properties. Fine tuning of the program code accelerated the rate of pKa prediction by a factor of about 3. We believe our prediction of ionization constants (pKa) is now the most comprehensive pKa model available.”