Bio-Rad to integrate Csiro molecular modelling

Technology uses Bayesian neural network technology to build in silico models for a wide range of end points such as Adme/Tox properties or biological activities

Bio-Rad Laboratories has announced plans to integrate Molsar (molecular structure activity relationships) molecular modelling technology into its KnowItAll informatics system as part of a proposed collaborative agreement with the Division of Molecular and Health Technologies of Australia's Commonwealth Scientific and Industrial Research Organisation (Csiro).

Molsar technology uses Bayesian neural network technology to build in silico models for a wide range of end points such as Adme/Tox properties or biological activities, allowing users to build their own predictive models using in-house experimental data and reference datasets pertaining to ligand moieties.

"Since medicinal chemists often work on compounds pertaining to very specific chemical classes, 'localized' models that are built using in-house data can be an optimal solution," said Bio-Rad's informatics division manager marketing and product manager Omoshile Clement.

Building models to predict the Adme/Tox profile of a drug compound is an area of increasing interest for drug discovery researchers due to the enormous potential of reducing discovery costs and time to market.

Building models "in house" is of interest to pharmaceutical and biotech companies since it provides the ability to create custom models without the risk of disclosing proprietary information to a third party.

"The integration of Csiro's Molsar technology into our KnowItAll platform will offer a wider array of in silico solutions," said Bio-Rad's informatics division general manager Gregory Banik.

"This partnership will mark a step forward in Bio-Rad's commitment to helping pharmaceutical researchers more effectively employ in silico technologies earlier on in the drug discovery process." The KnowItAll platform includes over 30 different pre-built models that can be used to predict properties such as water solubility, plasma protein binding, bioavailability, metabolism, mutagenicity and more.

Using in-house or proprietary datasets, the Molsar algorithm integrated within the KnowItAll platform will be able to be used to build, refine and/or optimize local models.

The KnowItAll platform can also combine local and global models into a single framework, allowing expanded model usage across broader chemistries.

"We are pleased to have an opportunity to share the benefits of this robust technology with drug discovery researchers around the world," said David Winkler, lead developer of the Molsar technology at Csiro.

Back to top