一种使用低复杂度方法进行多序列相似性计算的高效并行算法。
An efficient parallel algorithm for multiple sequence similarities calculation using a low complexity method.
作者信息
Marucci Evandro A, Zafalon Geraldo F D, Momente Julio C, Neves Leandro A, Valêncio Carlo R, Pinto Alex R, Cansian Adriano M, de Souza Rogeria C G, Shiyou Yang, Machado José M
机构信息
Department of Computer Science and Statistics, Sao Paulo State University, Rua Cristóvão Colombo 2265, 15054-000 São José do Rio Preto, SP, Brazil.
Department of Control Engineering and Automation, Federal University of Santa Catarina, Rua Pomerode 710, 89065-300 Blumenau, SC, Brazil.
出版信息
Biomed Res Int. 2014;2014:563016. doi: 10.1155/2014/563016. Epub 2014 Jul 22.
With the advance of genomic researches, the number of sequences involved in comparative methods has grown immensely. Among them, there are methods for similarities calculation, which are used by many bioinformatics applications. Due the huge amount of data, the union of low complexity methods with the use of parallel computing is becoming desirable. The k-mers counting is a very efficient method with good biological results. In this work, the development of a parallel algorithm for multiple sequence similarities calculation using the k-mers counting method is proposed. Tests show that the algorithm presents a very good scalability and a nearly linear speedup. For 14 nodes was obtained 12x speedup. This algorithm can be used in the parallelization of some multiple sequence alignment tools, such as MAFFT and MUSCLE.
随着基因组研究的进展，比较方法中涉及的序列数量大幅增加。其中，有许多生物信息学应用程序使用的相似性计算方法。由于数据量巨大，将低复杂度方法与并行计算结合使用变得很有必要。k-mer计数是一种非常有效的方法，具有良好的生物学结果。在这项工作中，提出了一种使用k-mer计数方法进行多序列相似性计算的并行算法。测试表明，该算法具有非常好的可扩展性和近乎线性的加速比。对于14个节点，获得了12倍的加速比。该算法可用于一些多序列比对工具（如MAFFT和MUSCLE）的并行化。
Zotero 插件