基于Spark云计算的生物基因多序列比对方法

A new approach for Biological Gene Multiple Sequence Alignment Based on Spark Cloud Computing

摘要:

在生物基因多序列比对过程中，早期的方法仅计算了单一的Spark集群参数，导致算法的并行效果较差。为此，设计了基于Spark云计算的生物基因多序列比对方法。基于获得的生物遗传序列数据，对其进行了优化，并通过计算不同序列间的匹配度，对生物基因多序列比对任务进行动态规划。利用Spark云计算技术，构建Spark集群，并对多个Spark集群的参数进行计算。利用多种生物基因序列之间的相似性与差异性来选择最佳的匹配路径，在此基础上，建立多个生物基因序列比对的并行计算模型，并对其进行求解，得到对应的多个序列对比对的并行算法。实验结果表明:该方法具有更好的并行性，能够有效提高多序列比对的性能。

Abstract:

In the process of multi sequence alignment of biological genes, parallel algorithms only calculate a single Spark cluster parameter, resulting in poor parallel performance of the algorithm. For this purpose, a parallel algorithm for multi sequence alignment of biological genes based on Spark cloud computing was designed. Based on the obtained biological genetic sequence data, it was optimized and the dynamic planning of the biological gene multi sequence alignment task was carried out by calculating the matching degree between different sequences. Utilize Spark cloud computing technology to build Spark clusters and calculate the parameters of multiple Spark clusters. By utilizing the similarities and differences between multiple biological gene sequences to select the optimal matching path, a parallel computing model for aligning multiple biological gene sequences is established and solved to obtain corresponding parallel algorithms for aligning multiple sequences. The experimental results show that the algorithm has better parallelism and can effectively improve the parallel performance of the algorithm.