Recent Publications  



Editor for the Following Journals

Design Drugs in Your Garage with SAnDReS  

Our goal here is to make available free computational tools for development of projects focused on in silico drug design. Researchers even with modest computer resources are still able to carry out projects for the development of potential new drugs using open source resources such as GemDockAutoDock, AutoDock VinaArgusLab, and SwissDock to mention few.

The big data generated in docking simulations can be analyzed using machine learning methods available in the program SAnDReS. An integrated approach with SAnDReS makes possible to identify potential new leads. The flowchart below brings the main steps to employ SAnDReS along your favorite docking program to carry out an integrate docking simulation, where you can bring together supervised machine learning techniques for regression and computational system biology to generate scoring functions tailored to your molecular system. For instance, if you are interested in identifying new potential inhibitors for the cyclin-dependent kinase, you may use your docking program native scoring functions and generate hundreds of polynomial scoring functions using SAnDReS, which may present better prediction power than native ones. 

One key point to evaluate docking performance is testing the ability to discriminate between active and decoy ligands. There over a hundred of datasets available in DUD-E that can be used to test this feature. In addition, DUD-E provides a free online system to generate decoy ligands for your active compounds, to do so, you need SMILES for the structures of active ligands.   

Flowchart for application of SAnDReS to analyze docking results and develop scoring functions. Grey boxes indicate task carried out by SAnDReS.

This concept follows the same idea proposed by Prof. Atul Butle who has described such initiatives focused on genetic data freely available on sites from the National Institute of Health. In this initiative, they have found a range of different ways of using this data to look for new drugs.