QCMI: A method for quantifying putative biotic associations of microbes at the community level

Abstract

A workflow has been compiled as "qcmi" R package-the quantifying community-level microbial interactions-to identify and quantify the putative biotic associations of microbes at the community level from ecological networks.

Figure 1

Conceptual framework used to assess putative biotic associations of microbial communities. Step 1. Construct ecological networks for sequencing‐based microbial communities. Step 2. Assign the assembly processes to each significantly paired ASVs. Step 3. Quantify the strength of putative biotic associations at the community level. Step 4. Calculate the effects of putative biotic associations of microbes on alpha and beta diversity. ASV, amplicon sequence variant; Env, environment; FS, forward selection; OLS, ordinary least squares; OTU, operational taxonomic unit; PC, personal computer; PCA, principal component analysis; PCNM, principal coordinates of neighbour matrices; SP, species.

Figure 2

Inference of putative biotic associations for the microbial communities under different habitats of freshwater, brackish, and saline wetlands. (A) Ecological networks were displayed in the Fruchterman–Reingold layout. Nodes indicated major taxa and links indicated ecological associations between nodes. (B) Assigned assembly processes to each significantly paired ASVs were shown via bar plots, including dispersal limitation, environmental filtering, the overlap of dispersal limitation and environmental filtering, and putative biotic associations. Colors represented different processes. (C) Comparisons in negative and positive associations of bacterial communities between original ecological networks (total) and filtered biotic networks (biotic). Student's t test was used to detect the differences. (D) The patterns of putative biotic associations within the microbial communities across low, medium, and high stress. ***Represented p < 0.001 and **represented p < 0.01. ASV, amplicon sequence variant.

Figure 3

Effects of putative biotic associations on microbial alpha‐ and beta‐diversity. (A) Random forest (RF) predicted the importance (percentage of increase of mean square error [MSE]) of drivers without and with biotic factors for species richness of bacterial communities. (B) Distance‐based redundancy analysis (dbRDA) ordination plotted the relationships between drivers without and with biotic factors and community dissimilarity of bacterial communities. The bottom bar showed the change in the change of explanation. Significance was expressed as ***p < 0.001; **p < 0.01; *p < 0.05. Lat, latitude; Long, longitude; MAP, mean annual precipitation; MAT, mean annual temperature; NA, negative associations; PA, positive associations; TC, total C; TN, total N; TK, total K; TP, total P.