Fig. 1

Comparison of the fecal microbiota in COVID-19 and non-COVID-19 patients. A, B The gut microbiota was compared between ventilated COVID-19 patients with antibiotics (COVID-19 group) to ventilated non-COVID-19 patients who were treated with antibiotics due to ventilation-associated pneumonia (pneumonia group) and those with neither pneumonia nor antibiotics (control group). A Chao1 and Shannon alpha diversity indexes in three groups at day 0 (inclusion) and day 7. The results of the Kruskal test and the pairwise Wilcoxon test are indicated within the graphs. B Composition of the fecal microbiota in three groups was represented in a non-metric multidimensional scaling (NMDS) plot of Bray–Curtis similarity. The results of pairwise group comparisons at day 0 and day 7 using PERMANOVA, and the effect of three covariates (COVID-19, antibiotics and timepoint) on microbiota composition of the three groups is indicated on the top left and bottom left, respectively. C, D Differential abundance analysis and distance-based redundancy analysis (db-RDA) in COVID-19 and non-COVID-19 patients. COVID-19 and non-COVID-19 pneumonia patients were compared to evaluate the COVID-19 effect. C Differentially abundant taxa identified by linear discriminant analysis effect size (LEfSe) at genus level. The results with p-value < 0.01 and effect size (log10) > 2.5 for each group are presented as a bar plot (**p < 0.01, ***p < 0.001). D Triplot demonstrating the relationship between the relative abundance of key taxa identified in LEfSe analysis and clinical variables, including COVID-19, ventilation, antibiotics and timepoint. Grey dots represent samples. The length and color of the arrows reflect the variance explained by the clinical variables and the significance of their effect on the gut microbiota composition, respectively. The short distance between the key taxa and the variables inversely indicates strong correlation. The RDA plot was based on Bray–Curtis distance and visualized with type 2 scaling