Experimental Data Analysis

Alpha diversity analyzes species diversity in a single sample and includes the Chao1 and Shannon indices (Jiang et al., 2015). The Chao1 index reflects the number of species in the sample microbiota (richness). In contrast, the Shannon index combines the number of species in the microbiota (richness) and the relative abundance of each species in the sample (evenness), expressed as species diversity in the microbiota.

Linear discriminant analysis effect size (LEfSe) (Fazlollahi et al., 2018) can be used to discover new biomarkers to identify the groups that best characterize each study group, and LDA analysis is primarily intended to find species that differ significantly in relative abundance between groups. LEfSe first detects differences in species abundance between groups using the non-parametric Kruskal-Wallis test to obtain species with significant differences; then, the Wilcoxon test was used to test the consistency of the differences between subgroups of the different groups in the first step, and finally, the LDA discriminant scores were used to estimate the magnitude of the effect of these differential species on intergroup distinctions. We considered taxa with LDA > 4 to be significant.

Sparse partial least squares discriminant analysis (sPLS-DA) (Hadj-Henni et al., 2021) aims to identify the key colonies that best distinguish between the two groups of samples, using the mixOmics package (Rohart et al., 2017) in the R language program for statistical analysis and graphing.