Construction of restriction maps using a genetic algorithm

Examines the design and performance of a genetic algorithm for double digest restriction mapping. Two objective functions were used to guide the algorithm which minimized the energy to the level of the true solution. In many instances, however, the algorithm was able to find arrangements which were even better, particularly for two enzymes and many fragments. An examination of the energy profile of the problem space for a typical problem reveals why this phenomenon occurs. More information from additional double digests improves discrimination but not performance.