Parallel computing for reconstructing physical maps of chromosomes

This study designs and implements the parallel algorithms with several optimization approaches including simulated annealing, large step Markov chains (LSMC), evolutionary programming, and genetic algorithms, for a physical mapping problem based on the maximum likelihood estimator model. The parallel algorithms are implemented using a combination of inter-process communication via message passing and shared memory multithreaded programming and have provided good performance. Genetic algorithms using a heuristic crossover operator yields better results in terms of both solution accuracy and performance compared to the simulated annealing, LSMC and evolutionary programming approaches.