Evolution is, in a sense, to resolve optimizaffon problems. Our neo-Darwinian algorithm based on the mechanics of inheritance and natural selection uses doublestranded DNA-type genetic information to resolve the "knapsack problem." The algorithm with asymmetrical mutations due to semiconservative DNA-type replication most effectively resolved the problem. Our results strongly suggest that disparity in mutations caused by the asymmetric machinery of DNA replication promotes evolution, in particular of diploid organisms with a high mutation rate, in a small population, and under strong selection pressure.Indispensable conditions for the evolution of an organism would be to increase adaptability, to win in competition, and to avoid species extermination. A major factor contributing to evolution seems to be spontaneous misreading of bases during DNA synthesis. Semiconservative replication of double-stranded DNA (dsDNA) is an asymmetric process-i.e., there is a leading and a lagging strand (1). This situation provides two possibilities for the occurrence of mutations: (i) a difference in frequency of strand-specific base misreading between the leading and lagging DNA strands (disparity model) and (ii) genetic algorithms, however, use a linear single-stranded genome (i.e., RNA-type), and mutations are inserted randomly during genome reproduction without semiconservative replication (4). We propose here another genetic algorithm (neo-Darwinian algorithm) to describe a disparity and a parity model of replication fidelity for dsDNA-type genome. We report the abilities of the algorithm to resolve a representative optimization problem in terms of model adaptability, competition between the two models, and extermination of models. Our neo-Darwinian algorithm mimics the semiconservative replication manner of dsDNA and its point mutations, which are inserted by base misreading when the genome replicates. In the present study, the so-called "knapsack problem" (4), which is rather simple but reflects the phenomenon of adaptation or the survival of the fittest organisms was selected to examine which model (disparity or parity) would obtain a better fitness score. Moreover, the effects of mutation rate, ploidy, population size, sexuality, and crossover were investigated with our neo-Darwinian algorithm.The results clearly indicated that the disparity model produced an overall higher fitness score than the parity model. Moreover, the disparity model was found to avoid extermination, even under extreme conditions. These observations suggest that in nature the presence of a difference in the fidelity between the leading and lagging DNA strands might contribute to the promotion of evolution. In other words, a driving force for the promotion of evolution might be inherent in the enzyme complexes that catalyze semiconservative replication of dsDNA.
NEO-DARVVINIAN APPROACH TO THE KNAPSACK OPTIMIZATION PROBLEMTo examine the fitness of the disparity or parity model, the neo-Darwinian algorithm was used to resolve the knapsack op...
