一种新宏基因组DNA片段物种属性约简分类方法的研究

由于宏基因组学的迅猛发展,导致可测序的宏基因组数量激增,可测DNA序列也在扩大,但绝大部分DNA序列种属未知,故生物信息学面临的重要课题是研究一种如何针对这些巨大量宏基因组DNA信息属性约简并正确分类的方法.提出采用RandWPSO结合粗糙集的优化算法,分别对宏基因组中的DNA片段在不同属、同属不同种情况下进行属性约简并分类,经实验证明能够进行有效约简可取得较高的分辩准确率.

A New Reduction and Classification Method on Metagenome DNA Fragments

The rapid development of the metagenomics has led to a surge in the number of sequenced macro genomes and the large number of DNA sequences that are predictable. However,most of the DNA sequences are unknown,so the important topic of bioinformatics is to study how to deal with the attribution reduction aiming at huge information of metagenome DNA fragments and find a method to classify them correctly. The algorithm of RandWPSO combined rough set was proposed to reduce and classify the DNA fragments in the metagenome ofdifferent genera but the same species,and it is proved that the effective reduction can be achieved by the experiment with a higher resolution accuracy.