Skip to main content
. Author manuscript; available in PMC: 2019 Jul 31.
Published in final edited form as: Cell Rep. 2019 Apr 23;27(4):1190–1204.e3. doi: 10.1016/j.celrep.2019.03.086

Figure 3. Microbiota of BALs with low CF pathogen abundance are more similar to paired scope washes than to paired mouth and throat samples.

Figure 3.

A. Total bacterial abundance in BAL samples, ordered from lowest to highest. Reference lines indicate mean and maximum scope pre-wash bacterial DNA loads. Ordering of BAL samples (n=102) is maintained in panels A-E. B. Relative abundance of CF pathogens (red bars) and non-conventional organisms (gray bars) in BAL samples. C and D. Morisita-Horn dissimilarity indices of BALs as compared to subject-paired mouth (C), throat (D), and instrument control (E) samples; 0 is maximum similarity, 1 is maximum dissimilarity. Dissimilarities were calculated using all identified bacterial OTUs. F. Mean relative abundance of taxa in all 102 BALs plotted as a function of their mean relative abundance in instrument controls. CF pathogen taxa (colored points) are labeled. The black line indicates the linear regression fit to non-conventional organism taxa (black points). G. Mean relative abundance of all non-conventional taxa in 34 BALs with low CF pathogen abundance (<164 CF pathogen genome copies/ml) plotted as a function of their mean relative abundance in instrument controls. CF pathogens were computationally removed for this analysis. The black line indicates the linear regression fit to the non-conventional organism OTUs. In F&G Pearson correlations are indicated for each fit (r and p-values). See also Figure S5.