Computational protocol: Structural modulation of the gut microbiota and the relationship with body weight: compared evaluation of liraglutide and saxagliptin treatment

Similar protocols

Protocol publication

[…] A raw sequence was assigned to samples according to the barcodes, and the valid sequence was further trimmed according to the following rules: each pyrosequencing read contained no more than two mismatches in the forward primer and less than three mismatches in the reverse primer, each pyrosequencing read had no ambiguous reads and less than 10 homologous reads, chimeric sequences were excluded, and all sequences were more than 200 bp in length. Trimmed sequences were clustered to operational taxonomic units (OTUs) defined by 97% similarity using the UPARSE method. The taxonomical assignment of OTUs was performed by alignment with the Bacterial SILVA database (SILVA version 115; http://www.arb-silva.de/documentation/release-115). OTUs that reached the 97% similarity level were used for diversity (Shannon, Simpson), richness (Chao, Ace), and rarefaction curve and Shannon-Wiener curve analyses using Mothur (version 1.30.1). Comparisons of diversity estimators and the metabolic indices between two groups were analysed by unpaired Student’s t-tests using SPSS 18.0. A heatmap figure was implemented by R packages gplots at the family and genus levels. An unweighted UniFrac distance metrics analysis was performed using the OTUs for each sample. A principal coordinate analysis (PCoA) was conducted according to the matrix of distance. A Metastats analysis was used to detect significant changes in relative abundance of the microbial taxa between groups. The linear discriminant analysis effect size (LEfSe) was used to select OTUs that exhibited significance (LDA value = 3.8) in the structural segregation among the grouping of samples. Correlations between OTU counts and the body weight of mice were determined using Spearman correlations. Given that the GLP-1 receptor agonist inevitably reduced food intake, a ridge regression analysis, which is thought to alleviate the effects of multicollinearity, was performed to explore the independent variables for the abundance of weight-relevant phylotypes. The ridge regression was conducted based on a ridge regression function implemented in SPSS 18.0 (Ridge, Regression, Analyze). Q-value for false discovery rate (FDR) control was used for the multiple testing correction (by R packages qvalue for ridge regression). […]

Pipeline specifications

Software tools UPARSE, mothur, gplots, UniFrac, Metastats, LEfSe
Applications Metagenomic sequencing analysis, 16S rRNA-seq analysis
Organisms Mus musculus
Chemicals Glucose