In many ways docking a candidate drug molecule to a Target structure is like fitting a key (candidate) to a lock (target).

Evaluation of the bonding energy between the Target structure and the candidate under scrutiny is repeated hundreds of thousands of times to identify the 3 dimensional orientation of the candidate that best fits the active site on the target, much like a key fits a lock. This process only requires the structures of the target and candidate molecules, which are very small, and lends itself very well to parallel and distributed computing applications.

The steps that are performed by (D2OL)™ during a docking are:

Download the Target structure if not already present locally on the computer.

Request work units that contain the structures of the candidates to be docked to the Target.

The base candidate structure is modified a number of times to create conformers - molecules that have the same atomic structure but have different orientation and 3 dimensional configuration that might "fit" better to the Target's active site.

Each conformer's docking energy is evaluated at different points of the Target's active site using a genetic algorithm and the lowest energy configuration is recorded.

After 10-30 different conformers have been created and evaluated, the results are sent back to (D2OL)™'s servers and a new candidate docking is initiated. The number of conformers tested depends on the Target that is being evaluated.

Post processing of the results takes place daily to rank and publish the best candidates for each Target on our site.

For more information please see our FAQ.

this page last reviewed October 1, 2004