Genome composition and phylogeny of microbes predict their co-occurrence in the environment

The genomic information of microbes is a major determinant of their phenotypic properties, yet it is largely unknown to what extent ecological associations between different species can be explained by their genome composition. To bridge this gap, this study introduces two new genome-wide pairwise measures of microbe-microbe interaction. The first (genome content similarity index) quantifies similarity in genome composition between two microbes, while the second (microbe-microbe functional association index) summarizes the topology of a protein functional association network built for a given pair of microbes and quantifies the fraction of network edges crossing organismal boundaries. These new indices are then used to predict co-occurrence between reference genomes from two 16S-based ecological datasets, accounting for phylogenetic relatedness of the taxa. Phylogenetic relatedness was found to be a strong predictor of ecological associations between microbes which explains about 10% of variance in co-occurrence data, but genome composition was found to be a strong predictor as well, it explains up to 4% the variance in co-occurrence when all genomic-based indices are used in combination, even after accounting for evolutionary relationships between the species. On their own, the metrics proposed here explain a larger proportion of variance than previously reported more complex methods that rely on metabolic network comparisons. In summary, results of this study indicate that microbial genomes do indeed contain detectable signal of organismal ecology, and the methods described in the paper can be used to improve mechanistic understanding of microbe-microbe interactions.微生物的基因组信息是决定其表型特性的核心因素，然而目前学界仍不甚明确：不同物种间的生态关联，在多大程度上可由其基因组组成加以解释。为填补这一研究空白，本研究提出两种全新的全基因组范围微生物间互作成对度量指标。其中第一种为基因组内容相似性指数（genome content similarity index），用于量化两株微生物间的基因组组成相似性；第二种为微生物间功能关联指数（microbe-microbe functional association index），可汇总针对特定微生物对构建的蛋白质功能关联网络拓扑结构，并量化跨物种边界的网络边占比。随后，研究人员利用这两种新指标，结合类群的系统发育相关性，对两份基于16S的生态数据集里的参考基因组间的共现关系进行预测。研究发现，系统发育相关性是微生物生态关联的强预测因子，可解释共现数据中约10%的变异；而基因组组成同样是强预测因子：当联合使用所有基于基因组的指标时，即便已纳入物种间的进化关系，其仍可解释共现数据中最高达4%的变异。单独使用时，本研究提出的这些指标所能解释的变异占比，要高于此前报道的依赖代谢网络比对的更为复杂的方法。综上，本研究结果表明，微生物基因组中确实包含可被检测到的物种生态特征信号，本文所述方法可用于深化对微生物间互作机制的理解。