Thanks for visiting Scientific Computing World.

You're trying to access an editorial feature that is only available to logged in, registered users of Scientific Computing World. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

Modelling technology wins prediction challenge

Share this on social media:

Optibrium's StarDrop Auto-Modeller provided the basis for a winning entry in a competition to build a model that can predict the environmental toxicity of new molecules, organised by ICANN'091, ENNS2 and CADASTER. The entry by Olga Obrezanova, principal scientist at Optibrium, came out joint winner in the challenge.

Ed Champness, CSO of Optibrium, said: 'I'd like to congratulate Olga and the StarDrop team on an excellent result in a difficult challenge that attracted 80 international participants. This demonstrates the power of the modern machine learning techniques in StarDrop to generate high quality, robust models, making predictive modelling more accessible to all research scientists.'

Obrezanova's recent work focused on combining the Gaussian Processes modelling method with StarDrop’s Auto-Modeller. This automatic framework gives both novice and expert users access to the tools needed to thoroughly analyse data and produce validated, predictive models to guide the design of new molecules in drug discovery. For non-experts, the Auto-Modeller can automatically generate models using multiple advanced modelling techniques and the built-in validation process rigorously tests and compares the resulting models to identify the most appropriate model to apply to new compounds. Alternatively, expert users can manually adjust modelling methods, data sets and descriptors to fine-tune the process.