Creation of a Calibration Set for Natural Product Virtual Target Screening

Jacob Wilson


Virtual screening (VS) is a computational method of drug design, which involves docking a collection of potentially efficacious molecules against a protein of interest which is known to be relevant in a disease. A few molecules which bind particularly well to the protein are identified and considered molecules of interest (MOI) and potential drugs. The problem with drugs discovered through this method is that they might have a high binding affinity for other proteins and those interactions may cause adverse effects. Virtual Target Screening (VTS) is a method which is used to identify the potential of an MOI to bind to other proteins. It is accomplished by virtually docking the MOI against a collection of proteins which are known to be prevalent and physiologically important in vivo. In order to determine the significance of an MOI’s interaction with each protein, the protein's average binding affinity for drug-like molecules must be determined by prescreening each protein against a collection of small synthetic molecules, which is often referred to as a calibration set. Since there is a new interest in the use of medicinal natural products, we have been working on adapting VTS to also work with natural products in addition to the synthetic molecules. Natural products, compared to synthetic molecules, can be larger, contain greater variety and complexity of aromatics, and can provide a greater representation of the diversity of chemical space. This project developed a natural product calibration set which was then run against several distinct proteins with known inhibitors. The ability of VTS to identify those inhibitors relative to the calibration set was tested to validate the calibration set. When such functionality is added to VTS, it will allow the identification of potential adverse interactions of natural product drug candidates and to repurpose existing natural product drugs.


Natural Products, Virtual Screening

Full Text: PDF


  • There are currently no refbacks.