engleski

Automating QSAR Applications & Visual Aids in Computer Aided Drug Design

As a systematic response to the challenges of data explosion in biomedical sciences, PLIVA R&D is in the process of establishing its informatics solution (project code-named PLIS) that will be capable of harnessing the power behind the tidal waves of information sweeping through the pharmaceutical industry. QSAR technology was recognized as one of the keys for better understanding relationships between biological activities and chemical structures, and its original implementation through a commercial MSI Life Science Suite was recently augmented by a set of intelligent agents that greatly simplify the interface and automate computational processes. In the front end, on the end-user's desktop, one can define a QSAR procedure by selecting: i) a set of classes of molecules ii) desired set of chemical descriptors iii) a user edited QSAR form or a predefined QSAR (i.e. Hansch equation) ; and finally iv) variables containing biological activities of interest In addition user can define criteria for running the background QSAR processes and a way of result notification. On the completion of the process intelligent agent will notify its owner, and continue to mine the database whenever the new data is present. Also, clustering and dimensional reduction techniques are used for additional representation of complex relationships among chemical properties and biological activity, with and emphasis on novel visualizations and an attempt to create a system for automated deduction of such relationships.

QSAR; chemical descriptors; novel visualisations

nije evidentirano