Fig. 2: The distribution of DPANN and CPR viromes mainly shaped by determinant factors.

a Associations of environmental factors and abundances based on correlation and random forest regression model. The circle size represents the variable importance (i.e., proportion of explained variability), and the color gradient in the heatmap denotes Pearson’s correlation coefficients. The bar chart shows the explanation of the response variable by the best model. C-host, the abundance of CPR. C-virus, the abundance of CPR virus. D-host, the abundance of DPANN. D-virus, the abundance of DPANN virus. b Distribution of sulfate (g/kg), Fe (g/kg), Pb (mg/kg), EC (mS/cm), and pH levels across MRT-based groups (A-F). For each box plot, the center line represents the median, the box limits represent the 25th and 75th percentiles (IQR), and the whiskers extend to 1.5×IQR. All individual data points are overlaid (Group A: n = 39; Group B: n = 34; Group C: n = 3; Group D: n = 5; Group E: n = 2; Group F: n = 6). Sulfate, Fe, and Pb concentrations are displayed on a log2 scale. c the relative abundance of CPR and DPANN in different groups. Colors were used to represent different taxa. d Viral composition in different groups. e, f NMDS results of CPR (e) and DPANN (f) viral structure colored by MRT-derived groups. Boxplots show distributions of different groups in NMDS1 and NMDS2. For each box plot, the center line represents the median, the box limits represent the 25th and 75th percentiles (IQR), and the whiskers extend to 1.5×IQR. Group differences were tested using PERMANOVA (adonis) and ANOSIM with 999 permutations based on Bray-Curtis distances.