中文摘要：

随着基因测序技术快速发展，产生了大量海洋微生物数据，使未培养（难培养）海洋微生物研究成为可能．常用的统计学方法和统计分析软件，无法深度挖掘隐含在这些大量数据中的海洋微生物作用模式及种群组成多样性．定义了边重要性测度及节点重要性测度，确定社团核心节点，基于节点．社团置信度，提出整合边、节点信息网络社团挖掘算法（CDIEV），并对春、夏、秋、冬4季节下的海洋微生物作用模式进行了研究．首先采用ESPRIT算法将16SrRNA基因序列聚类成微生物操作分类单元（OTU），基于Spearman相关性及P值分别构建春、夏、秋、冬4季微生物作用网络．CDIEV算法仿真结果及网络拓扑参数分析表明：春、夏、秋、冬海洋微生物作用网络具有复杂网络“小世界”和“无尺度”特性；4个季节下的海洋微生物作用模式存在一定的差异；CDIEV算法可有效挖掘网络社团，但挖掘结果受添加阈值影响．

英文摘要：

With the development of high-throughput and low-cost sequencing technology, a large amount of marine microbial sequences are generated. So, it is possible to research more uncultivated marine microbes. The interaction patterns of marine microbial species and marine microbial diversity hidden in these large amount sequences can not be deeply mined with the conventional statistical methods and analytical software. In this paper, we defined the importance measures of network edges and vertex, identified the key vertices of the community, then defined the confidence of one vertices belonging to the community. In the end, we proposed the community detection algorithm by integrating the edge and vertices information in complex network （CDIEV）, and also applied it to uncover the seasonal marine microbial interaction patterns. After the marine microbial 16S RNA sequences clustered into operational taxonomic units （OTUs） at 99% sequence similarity with ESPRIT algorithm, the four seasonal （spring, summer, autumn, winter） marine microbial interaction networks based on the spearman correlation and P values were constructed. The experimental results of CDIEV and analysis of network topological parameters show that the four seasonal marine microbial interaction networks have the characters of small-world and scale-free of complex network; the marine microbial interaction patterns have some difference among the four seasonal marine microbial interaction networks; CDIEV algorithm can effectively mine the network community, but the selection of threshold value fl affects the community detection results.