A 3-D QSAR MODELS DATABASE for Virtual Screening
Substance libraries, composed of million of compounds, are tested in the pharmaceutical industry with robotic systems. Furthermore, the follow-up of the “hits” is often so expensive that essentially only large companies can use this method.
Virtual screening helps to decide which compounds to screen, which libraries to synthesis and which compounds to purchase from an external supplier reducing the overall cost associated to the discovery and development of new drugs.
Main classes of virtual screening methods are:
• Similarity search (ligand-based virtual screening)
• Identify a common 3-D pharmacophore, then do a 3-D database search
• Train a machine learning technique
• Protein-ligand docking
Most of these methods are computationally intensive and complex.
3-D QSAR methods are nowadays used widely in drug design, since they are computationally not demanding and afford fast generation of QSARs from which the biological activity of molecules can be predicted.
We have developed a software system for automatically score, rank and/or filter a set of structures using pre-built 3-D QSAR models.
A Web interface incorporating a molecular drawing interface enables the users to process their own molecules by drawing or uploading them to the server and selecting the target for the virtual screening and biological activity prediction.
Multiple targets screening are allowed, making the RCMD 3-D QSAR Server a very interesting tool for virtual high throughput screening.
The system consists of a library of pre-built 3-D QSAR models. This database is constantly updated with published or newly developed 3-D QSAR models.