Mining correlation patterns of taxa, pathways and environmental factors with an improved weighted network community detection algorithm

With the development of high-throughput and low-cost sequencing technology, a large amount of marine microbial sequences is generated. So, it is possible to research more uncultivated marine microbes. Generally, the functional capability and taxa structure are highly related with environment factors in microbial communities, which are hidden in these large amount sequences. However, most works used the canonical correlation analysis (CCA) method to research the correlative relationship among taxa, pathways and environmental factors. CCA is difficult to find which environmental factors are the major determinants of some special taxa and pathway. In this paper, we integrated 14 ocean metagenomes with geographical, meteorological and geophysicochemical data to construct the correlative weighted networks with Spearman correlation. By using an improved weighted network community detection algorithm, named as IWNCD, we find some special correlation patterns among taxa, pathways and environmental factors. Analysis of these patterns shows that the climatic factors such as temperature, sunlight, and correlated CO₂, and the nutrients such as chlorophyII and primary production are the main determining factors of the functional community composition; The growth and development of some special taxa are dependent on some main environmental factors such as sunlight, temperature, CO₂, primary production, dissolved oxygen, dissolved silicate; In addition, sampling sites more similar in geographic location have a greater tendency to be closer together based on their metabolic pathways.