Bio-Rad Announces New Software Technology for In-Silico ADME | Tox Modeling

PHILADELPHIA, PA, August 12, 2003 – Bio-Rad Laboratories, Inc. (AMEX: BIO and BIO.B), a multinational manufacturer and distributor of life science research products and clinical diagnostics, announced today that their Informatics Division has introduced a new ADME/Tox Consensus Modeling Environment for use in in silico ADME/Tox screening. The new technology has been integrated into the Company’s KnowItAll® Informatics System, used for ADME/Tox prediction, spectroscopy data management, and cheminformatics.

The new Consensus Modeling Environment is a software technology that optimizes the combined results of multiple in silico ADME/Tox models.  The Consensus Modeling Environment has the capacity for greater accuracy than any of the individual models. As Michelle D’Sousa, Ph.D., Medicinal Chemistry Software Product Manager for Bio-Rad’s Informatics Division, remarked, “By taking a weighted average of individual continuous variable models or by utilizing multiple classification models, we can achieve greater accuracy than any of the models alone.”

Initially, the Consensus Modeling Environment will be applied to two types of discrete prediction models—log P (octanol/water partition coefficient) and water solubility—and one type of classification model—mutagenicity. Bio-Rad Informatics has entered into agreements with developers of these technologies to deploy multiple models in the Consensus Modeling Environment. Bio-Rad eventually plans to extend this technology into other predictive models of interest to ADME/Tox investigators.

The Consensus Modeling Environment employs an application called “ValidateIt™” to enable the user to “train” multiple continuous variable models against a set of chemical structures with known values. This training process calculates the best weighting of the individual models to most closely match the known values and automatically calculates an N-fold cross-validation statistic. For consensus models created from continuous variable models, ValidateIt also provides the user with scatter plots for each model (including the Consensus Model) and tables that detail the weighting and the statistical error. Any number of trained consensus models can be saved and applied to structures of interest at any time.  For consensus models created from classification models, ValidateIt automatically calculates the concordance, false positives, false negatives, sensitivity, specificity, and accuracy of the consensus model and each individual model and allows the corresponding structures for each subset to be displayed.

Gregory M. Banik, Ph.D., Bio-Rad, Informatics Division General Manager, commented, “Our goal  to provide a single, integrated ADME/Tox Consensus Modeling Environment that covers a wide range of applications has been achieved with the new ValidateIt application. In silico ADME/Tox modeling is an area of growing interest in Drug Discovery, and by integrating many applications in Bio-Rad’s Consensus Modeling Environment, we provide the user with the best possible prediction accuracy with integrated validation capability, thus eliminating the risk of choosing and purchasing poor-quality, unvalidated models.”  He continued, “We invite other ADME/Tox technology providers to add their models to the Bio-Rad’s KnowItAll Consensus Modeling Environment.”

Bio-Rad’s Informatics Division specializes in state-of-the-art software and database solutions for the pharmaceutical, biotech, and chemical industries. Bio-Rad is the leading publisher of fully verified spectral databases, cheminformatics, spectroscopy software, and decision support systems for drug discovery. The company’s award-winning KnowItAll Informatics System offers a fully integrated environment with flexible, expandable software and database solutions for spectroscopy (MS, NMR, IR, Raman, spectral data management), cheminformatics, and ADME/Tox evaluation.

Bio-Rad Laboratories, Inc. (www.bio-rad.com) is a multinational manufacturer and distributor of life science research products and clinical diagnostics. It is based in Hercules, California, and serves more than 70,000 research and industry customers worldwide through a network of more than 30 wholly owned subsidiary offices.

Various statements made within this press release may constitute “forward-looking statements” for purposes of the Securities and Exchange Commission’s “safe harbor” provisions under the Private Securities Litigation Reform Act of 1995 and Rule 3b-6 under the Securities Exchange Act of 1934. The forward-looking statements contained herein involve risks and uncertainties that could cause results to differ materially from the Company’s expectations.

For more information, contact:
Name: Deborah Kernan
Phone: +1 215 349 7366
E-mail: deborah_kernan@bio-rad.com

For additional information on Bio-Rad Laboratories, Informatics Division, visit www.informatics.bio-rad.com or call 1-888-5-BIO-RAD in the U.S. or in Europe, call +44 (0) 20 8328 2555 or free phone 00800-78945000.