In drug discovery research, a drug candidate is typically a small molecule (ligand) that can make a strong binding to the drug target. The target is typically a protein that is either a receptor responsible for transmitting signals to a cell, or an enzyme that enhances the rate of chemical reactions. A drug candidate can interfere with these biological processes incurred by the target molecule, if it can bind with high affinity at specific binding sites. Therefore, scientists usually search available libraries of ligands for putative drug candidates; this process is called screening or docking. Biochemical screening, however, requires expensive lab equipment and takes considerable time. Alternatively, the initial phase of drug discovery nowadays involves virtual screening software tools that simulate binding of ligands and calculate binding affinities. Only drug candidates with high affinities (resulted from simulation) will be further studied in the wet lab.
Life Sciences, Biochemistry
AutoDock Vina is one of the most advanced software suites in molecular modeling simulation. It is especially effective for Protein-ligand docking. AutoDock Vina is available under the Apache license. AutoDock Vina achieves an approximately two orders of magnitude speed-up compared to the AutoDock 4 molecular docking software, while also significantly improving the accuracy of the binding mode predictions. Further speed-up is achieved from parallelism, by using multithreading on multi-core machines. AutoDock Vina automatically calculates the grid maps and clusters the grid maps and clusters the results in a way transparent to the user.
The workflow (shown below) starts with the Prepare component, which takes as input, the receptor file (on port 0), the configuration file (on port 1) and a list of ligands as a zipped file (on port 2). The Prepare component splits the ligands into distinct groups and passes them to the AdVina component, along with the receptor and configuration files. Several instances of the AdVina component are executed to process these ligand groups in parallel. Finally, the Collect component collects all the outputs, creates a sorted list of the binding energies, and packs the outputs into a single compressed file.
The workflow dynamically maximizes parallelization granularity based on two restrictions: a minimum number of ligands per job is needed to minimize the overhead versus computation time; and, a maximum splitting factor is used to lower the overhead on gUSE.
A particular molecule related to the development of atherosclerosis has been studied for many years at the AMC. This molecule seems to be a natural mechanism to protect blood vessels from plaque development (=slows down the development of plaque).
There is interest in controlling the activation of this molecule with drugs, however to but this is not easy to do. The general understanding today is that it cannot be artificially controlled, however the researchers of the AMC still have hope.
Virtual screening for compounds that could activate this molecule is even more challenging because there is no obvious place to look at. Lots of docking simulations are needed. It is like looking for a needle in a haystack – without knowing upfront that the needle is actually there. Large screening experiments have been started at the AMC from the docking gateway to investigate means to activate this molecule. It will take many years still to find out whether this molecule can be activated with drugs.
PS: the name of the molecule is omitted here to protect intellectual rights of the AMC researchers.
- MM Jaghoori, et al. A Grid-Enabled Virtual Screening Gateway. In the 6th International Workshop on Science Gateways (IWSG 2014), pages 24-29.