Additional file 1 of Sputum bacterial load and bacterial composition correlate with lung function and are altered by long-term azithromycin treatment in children with HIV-associated chronic lung disease

Additional file 1: Figure S1. A bar plot of the taxa and their relative abundance of the extraction and sequencing mock controls compared to manufacturer profiles. Figure S2. A scatterplot showing the correlation between samples repeated within a run (WR, n = 74) and between runs (BR, n=28). Figure S3. A scatterplot showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) 16S copies vs final number of reads (A1 and A2), Shannon alpha diversity index (B1 and B2) and age of participant in years (C1 and C2). Figure S4. Ordination plots of showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) coloured by their 16S copies. Figure S5. Ordination plots showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) coloured by their number of reads. Figure S6. Ordination plots showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) coloured by the age of the participant. Figure S7. Rarefaction curves showing number of ASVs detected and 16S copies of samples. Figure S8. Rarefaction curves showing number of ASVs detected and number of reads of samples. Figure S9. Bar plot showing the profiles of biological samples with <100 16S copies (n=2) in comparison to Primestores profiles (n=43). Figure S10. Bar plot showing the profiles of biological samples with >100 to <1000 16S copies (n=10) in comparison to Primestores profiles (n=43). Figure S11. Ordination plots showing the profiles of a subset of biological samples with low 16S copies and the negative controls. Figure S12. Ordination plots showing the profiles of a subset of biological samples with low reads and the negative controls. Figure S13. Ordination plots showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) coloured by the run in which the sample was processed. Figure S14. Ordination plots showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) coloured by the country of sampling. Figure S15. Ordination plots showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) coloured by visit. Figure S16. Ordination plots showing the spread of biological samples (n=960) and the negative controls (primestore, n=43) coloured by the age at sampling. Figure S17. Output from decontamination analysis using the DECONTAM R package. Figure S18. Boxplot of Shannon alpha diversity index between trial arms at each visit (A) and between study visits in AZM (B) and Placebo (C) arms. Figure S19. Violin boxplot comparing two beta diversity metrics between samples collected from participants in the AZM arms at baseline and 48 weeks, 48 and 72 weeks and baseline and 72 weeks. Figure S20. Violin boxplot comparing two beta diversity metrics between samples collected from participants in the Placebo arms at baseline and 48 weeks, 48 and 72 weeks and baseline and 72 weeks. Figure S21. Principal Coordinates Analysis of Atchison (A) and Bray-Curtis (B) [on unrarefied ASV counts] distance matrixes between trial arms at each visit. Figure S22. Barplot of the relative abundances of the top 10 most prevalent phyla in all samples. Figure S23. Barplot of the relative abundances of the top 12 most prevalent genera in all samples. Figure S24. Heatmap displaying the q values of the genera detected as differentially abundant between AZM and placebo arms at 48 weeks by 10 statistical methods. Figure S25. Heatmap displaying the q values of the genera detected as differentially abundant within the AZM arm between baseline and 48-week samples by 10 methods. Figure S26. Heatmap displaying the q values of the genera detected as differentially abundant within the AZM arm between 48- and 72-week samples by 10 methods. Table S1. The taxonomy of the ASVs in the extraction and sequencing control. Table S2. List of 70 ASVs detected by the DECONTAM R package as potential contaminants based on comparison between biological samples and negative controls. Table S3. The association between bacterial load (16S rRNA copies) and selected variables using linear mixed effects modelling. Table S4. The association between Shannon diversity indices and selected variables using linear mixed effects modelling. Table S5. Results of differential abundance testing of bacterial taxa from AZM and Placebo samples from 48 weeks using 10 methods. Table S6. Results of differential abundance testing of bacterial taxa from AZM and Placebo samples from 72 weeks using DESeq2. Table S7. Results of differential abundance testing of bacterial taxa from AZM and Placebo samples from 72 weeks using Ancom-II. Table S8. Results of differential abundance testing of bacterial taxa from samples from the AZM arm at baseline and 48 72 weeks using 10 methods. Table S9. Results of differential abundance testing of bacterial taxa from samples from the AZM arm at 48 and 72 weeks using 10 methods. Table S10. Results of differential abundance testing of bacterial taxa from Placebo samples from 48 and 72 weeks using DESeq2. Table S11. Results of differential abundance testing of bacterial taxa from Placebo samples from baseline and 72 weeks using DESeq2. Table S12. Contributions of top genera to overall dissimilarity between AZM and Placebo arms at 48 weeks and, within the AZM arm, between Baseline and 48-week samples- SIMPER analysis. Table S13. Univariate linear regression analysis of within-participant Aitchison distance (outcome) and within-participant change in lung function metrics (FVCz and FEV1z) between visits.