微生物多样性（扩增子/16S rDNA测序）—功能预测分析方法描述

​一、代谢、功能预测分析内容及意义

    根据已知的微生物基因组数据，对菌群组成的测序数据（典型的如16S rRNA基因的测序结果）进行菌群代谢功能的预测，

    意义：从而把物种的“身份”和它们的“功能”对应起来。根据菌群代谢功能预测结果，我们一方面能一窥菌群功能谱的概貌，发挥菌群多样性组成谱测序性价比高的优势；另一方面也能帮助指导后续宏基因组De novo鸟枪法测序的实验设计，更合理地筛选用于后续研究的样本。

a) PICRUSt

 通过将现有的16S rRNA基因测序数据与代谢功能已知的微生物参考基因组数据库相对比，从而实现对细菌和古菌代谢功能的预测；预测过程中还考虑了不同物种16S rRNA基因拷贝数的差异，并对原始数据中的物种丰度数据进行校正，使预测结果更准确可靠。

 局限性：已知菌种、已知功能、参考数据库是否充足、基因转移等因素造成误差、          只能用Greengene数据库。

b) 其他软件：Tax4Fun等

二、功能、代谢预测分析在论文中的描述 

a) PICRUSt

方法描述示例：Functional predictions using PICRUSt

PICRUSt is a bioinformatics tool that uses marker genes to predict the functional content of microorganism. In this study, this method was employed to predict the potential functions of each sample based on 16S rRNA sequencing data. We used the KEGG database[和/或 COG database等] and performed closed reference OTU picking using the sampled reads against a Greengenes reference taxonomy (Greengenes database Version). The 16S copy number was then normalized, molecular functions were predicted and final data were summarized into KEGG pathways. The differences in predicted molecular functions of bacterial communities among four groups were shown by PCA (principal component analysis) using R “vegan” package. ANOSIM was used to test whether the dissimilarity of gene abundance among the four groups was significant.

结果描述示例：Functional predictions with PICRUSt

We used PICRUSt to predict changes in microbial functions that might be associated with changes in OTU abundance detected via 16S sequencing. The PICRUSt approach has been proven useful to predict genomes of organisms in environmental samples and may offer insights on the potential functions of goose gut microbiota. In this study, the chosen reference OTUs were used to match the KEGG database to predict microbial functions. Using this method, our study inferred 41 gene families in the faecal samples (Fig. 6a). We also performed PCA of the relative abundance of KEGG pathways to reveal the clustering of samples (Fig. 6b). The histogram and PCA plot both revealed that the potential functions of the microbiota of the four groups were similar (ANOSIM, p>0.05).

分析示意图

The majority of the 41 gene families belonged to membrane transport (11.77%), carbohydrate metabolism(9.30%), amino acid metabolism (9.24%), replication and repair (8.93%), energy metabolism (7.34%), translation (5.88%), poorly characterized (4.63%), metabolism of cofactors and vitamins (4.59%), nucleotide metabolism (3.72%), and cellular processes and signalling (3.49%). The abundances of most of the gene families of GWFG-SJL differed significantly from the abundances of the other three groups in PYL, with 33 genes from GWFG-PYL, 36 from BG-PYL and 25 from. However, the relative abundance of each gene family was more similar among the three data sets in PYL. Among the ten dominant gene families noted above, the abundance of genes related to energy metabolism, carbohydrate metabolism, metabolism of cofactors and vitamins, amino acid metabolism and poorly characterized families were significantly higher in the GWFG-SJL samples than in all other data sets (Fig. 6c).