An efficient clash detection method for molecular structures applications

Molecular structures applications are usually computationally intensive. Problems like protein docking, ab-initio protein folding and even some visualization software that verify structure correctness, need to constantly determine if two atoms in the structure collide. This problem is usually referred as clash detection. If for the ab-initio and docking problems the cost of these computations determines the number of configurations that can be searched, for the visualization program the cost determines the application response time. Either way, a fast clash detection method can greatly improve the application effectiveness. This work focus mainly in the ab-initio protein folding problem. Several protein folding methods avoid collisions by using energy functions that consider the Van der Walls forces. For other methods, which apply less expensive scoring functions, a faster clash detection algorithm may greatly increase the overall time efficiency. This will allow for more structures to be considered and thus improve the chances of finding the right structure. We propose an efficient data structure with which we can achieve constant time clash detection, independently of the size of the protein. We compare the proposed data structure with one of the best known general data structures for this type of problems (range queries) and with the naive approach. The results show that the proposed data structure surpasses the other techniques for any protein size. The proposed data structure takes near half the time of the general data structure and close to a fifth of the time of the naive approach for the larger proteins.