Abstract: Given a set of weighted aligned DNA fragments of an individual，the haplotype assembly problem Weighted Minimum Letter Flips（WMLF） is to infer a pair of haplotypes of the individual by flipping some SNPs with the sum of weights of the flipped SNPs minimized.WMLF is NP-hard，and there is no practical algorithm to search optimal solution for the problem.Based on the characters of DNA fragments，this paper introduces a genetic algorithm.For the practical fragment data，the algorithm can solve the WMLF problem efficiently even if the data are larger，and it is scalable and applicable in practice.

摘要： 单体型组装加权最小字符翻转（WMLF）问题指定个体联配的加权DNA片断数据，翻转权值和最小的SNP位点以推测出该个体的一对单体型。该问题是NP-难的，至今尚无实用的搜索寻优算法。根据DNA测序片段数据的特点提出了一种遗传算法。对于实际的生物实验数据，即使数据很大，该算法也可以在较短的时间得到WMLF问题的满意解，具有良好的可扩展性和较高的实用价值。