Molecular Docking Game

Molecular docking is an important problem in biology for the study of immune systems, allergies, and many more functions. Automated methods for solving molecular docking can be computationally expensive, instead, we use an approach to take advantage of human intuition. To accomplish this, we develop a molecular docking game that could be expanded to a crowdsourcing application. The molecular docking game records all states of ligand proteins during gameplay. Single states can be scored and selected as potentially docked states, while all the states can be combined incrementally into a roadmap which lets us perform motion planning between two ligand states.

The haptic device provides 3 degrees of freedom force feedback to the user as they move the ligand. This force feedback is based on the gradient of the potential energy between the atoms in the ligand and receptor protein. A player using a haptic device can be guided by this feedback into possible docking sites for the ligand. Each molecular docking game session, the user finds the best possible docking site they can.

The score shown back to the user is based off of the potential energy and reflects the quality of the possible docking site. All the ligand states in between these possible sites are also recorded to a customizable amount of precision. These states can be combined into sets from multiple players and sessions for roadmap construction. Each edge in the roadmap represents a transition between the two ligand states and is weighted by the potential energy barrier. Finding the shortest weighted path through this roadmap is useful for predicting possible ligand motion paths around the receptor protein.

Ligand states recorded from gameplay are built into a roadmap

Score Feedback When Finding Lower Energy States

The Ligand Fits in Many Locations on Receptor

High Score List Can Show Relative Performance