Molegro brings GPU acceleration to computational drug discovery

AARHUS, DENMARK: Molegro has released a new version of Molegro Virtual Docker, an integrated platform for predicting protein-ligand interactions. Molegro Virtual Docker enables chemists to computationally predict and visualize how small molecules interact with a protein receptor, making it possible to understand and rationally optimize potential drug candidates.

Even if the 3D protein structure is not known, it is still possible to make predictions based on similarity to known binders, or to train regression models based on molecular descriptors - all inside an integrated graphical environment.

In Molegro Virtual Docker 5, it is now possible to take advantage of Graphics Processing Units (GPU's) to greatly accelerate the docking simulations. Though GPU's were originally designed for pure visualization purposes, their computational power has now outpaced traditional CPU's, making them an attractive target for demanding scientific computations.

Using the GPU acceleration in Molegro Virtual Docker, it is now possible to predict how a molecule will bind to a target receptor in less than one second on a single machine - making it more than twenty times faster than a pure CPU-based approach. Screening large compound databases may now be done using a fraction of the hardware needed when doing conventional CPU based simulations, and becomes a viable task without requiring a large cluster of machines. The GPU acceleration features in Molegro Virtual Docker require a NVIDIA GeForce graphics card or NVIDIA Tesla GPU with CUDA support.