Extended Data Fig. 7: The airway microbiome healthy index.

a) AUC and accuracy of AMHI calculated based on the functional features (KOs, ARGs and VFs) in distinguishing airway health and disease status. b) Violin plots showing a significant decrease of AMHI (using amplicon-based bacterial and fungal genera) in disease over healthy individuals across all 6 districts (Wilcoxon rank-sum test, two-sided). The number of individuals is indicated in the parenthesis. c) Violin plots showing the association of AMHI with airway symptoms among airway healthy individuals only. For respiratory symptoms, the P-values were obtained in comparison with the no symptom group using Wilcoxon rank-sum test (two-sided). The number of individuals in each group is indicated in the parenthesis. Exact P-values (top to bottom): 0.0727, 0.538, 0.0330, 0.316, and 0.00388. d) The interaction effects of AMHI with biofuel exposure, second-hand smoking, and occupational pollution on their effects on the high respiratory symptom burden (CAT > = 10). Shown are the estimate and P-value of the interaction term in the general linear model, and the increased odds of having a high symptom burden in exposure to occupational pollution with one unit decrement of AMHI (Δ odds). e) The top KMs, ARGs and VFs correlated with AMHI. For display purpose, KMs with FDR < 0.005 in association with AMHI are shown. ARGs and VFs with FDR < 0.05 are shown. f) Violin plots showing an extrapolation of AMHI to 5 external sputum microbiome datasets on healthy smokers and non-smokers. For each dataset, the relative AMHI scores in smokers normalized to the average and standard deviation of non-smokers are shown in the violin plot. P-value was obtained using Wilcoxon rank-sum test (two-sided). Exact P-values (left to right): 0.0436, 0.0576, 0.00354, 0.235, and 0.00856. The numbers of smokers and non-smokers are indicated for each dataset. For the boxplots within the violin plots, the central line indicates the median. The lower and upper hinges indicate the first and third quartiles. The lower and upper whiskers extend from the hinge to the smallest and largest values no further than 1.5 * inter-quartile range from the hinge. *** P < 0.001, ** P < 0.01, * P < 0.05, + P < 0.1.