微生物生态研究中BIOLOG方法数据分析及R语言实现

微生物生态研究中，对微生物群落结构、群落特征以及其与环境因素的关系的揭示，一直受到广泛关注；适当的数据分析方法有助于更清晰地刻画微生物群落结构特征，明确其与环境因素的关系。结合实例，对微生物生态研究中基于BIOLOG微平板技术的数据分析方法进行梳理，分别介绍数据读取整理、特征指数计算、非限制性排序、限制性排序、聚类分析、环境向量拟合、蒙特尔检验等常用数据操作及生态分析方法；针对不同方法结论，结合研究目标和生态理论给出具有统计学意义的解释，并评价不同方法特点及适用场景；分析过程以R语言实现，并提供全部代码。结果表明，BIOLOG方法产生数据能从多个角度表征微生物群落功能特征，并结合环境指标梯度进行分析；但BIOLOG数据可能不满足正态性分布，在基于正态分布的分析前应提前进行检验；排序分析时应慎用主成分分析，可优先采用其他基于距离矩阵的排序方法；R语言能够简化BIOLOG数据读取及操作，易于完成各类统计分析。本研究能够对微生物生态研究者科学选择应用统计分析方法、提高数据处理效率提供直接参考。

Data analysis and R demonstration for application of BIOLOG microarrays technique in microbial ecology study

In the study of microbial ecology, there have been increasing concerns about community patterns and their relationship with environmental factors. Selecting appropriate statistical methods not only provides a better understanding of the microbial community structure and pattern but also clarifies the effects of environmental factors on the microbial community. Based on practical examples, this paper summarized the statistical methods for BIOLOG data in microbial ecology study, including data input, indicator indices calculation, unconstrained ordination, constrained ordination, clustering analysis, environmental factor fitting, Mantel test, and so on. The statistical results were discussed based on study purposes and ecological theories, and the applicability of those methods was also evaluated. All analysis was conducted with R programming. The results show that BIOLOG microarrays techniques effectively reveal the physiological profiles of the microbial community. Moreover, this technique can be appropriately used to analyze the relationship between microbial profiles and their environmental factors. Here, we should recognize that the pre-inspections are required, as BIOLOG data maybe not in a normal distribution. For ordination analysis, some other applications of distance-based ordination methods are safer than Principal Component Analysis (PCA), which should be given a second thought. The priority should be given to the ordination analysis based on the distance matrix rather than PCA. Furthermore, R programming facilitates the BIOLOG data manipulation. The study can help researchers choose proper analysis in microbial ecology study and increase the efficiency of data manipulation.