图1 玉米叶际微生物在4种互作关系网络下的中心节点微生物。在4个不同的共现网络中, 中心节点微生物是基于特征值度和接近中心性进行划分的, 每个网络的右侧显示分类操作单元(OTU)的这两个值, 进行中心节点微生物的筛选, 并把中心节点微生物在网络中用不同的数字进行表示。OTU对应的物种组成见表1, 最终使用Gephi软件对4种互作关系进行可视化。

Fig. 1 Hub microbes of maize phyllosphere microorganisms under the four different co-occurrence networks. In four different co-occurrence networks, hub microbes are divided based on degree centrality and closeness centrality, the right side of each network shows these two values of the operational taxonomic unit (OTU). The hub microbes are represented by different numbers in the network, the species composition corresponding to OTU is shown in Table 1, and finally the four interaction relationships are visualized using Gephi software.

图2 玉米叶际微生物组4个共现网络的6种网络特征值。横坐标从左到右依次代表6种网络特征值, 纵坐标代表230个不同基因型玉米样本, 展现不同个体之间、不同共现网络之间网络特征值的差异。图右边0-1.0的颜色变化代表差异的大小, 0代表没有差异, 1.0代表差异最大。六种网络特征值分别是: 连通性(Con)、接近中心度(C(u))、中介中心度(B(u))、偏心率(E(u))、特征向量(G(u))、网页排名(P(u))。

Fig. 2 Six network indices of four co-occurrence networks of maize-phyllospheric microbiome. As shown in the figures above, the abscissa represents six network feature values from left to right, and the ordinate represents 230 different genotype maize samples, showing the difference in network feature values between different individuals and between different co-occurring networks. The color change of 0-1.0 on the right side of the figure represents the size of the difference, 0 represents no difference, and 1.0 represents the largest difference. The six types of network eigenvalues are: connectivity (Con), closeness (C(u)), betweenness (B(u)), eccentricity (E(u)), eigenvector (G(u)) and pagerank (P(u)).

图3 四种相互作用关系下广义线性模型(GLM)、混合线性模型(MLM)、FarmCPU的比较。通过网络作图得到GLM、MLM、FarmCPU模型并用分位图进行展示。四种互作关系下每个网络特征值进行3种模型的比较, 最接近中间红色线的为最优模型。

Fig. 3 Comparison of General Linear Model (GLM), Mixed Linear Model (MLM), Fixed and Random Model Circulating Probability Unification (FarmCPU) under four interaction relationships. GLM, MLM, and FarmCPU models are obtained through network mapping and displayed with Quantile-Quantile diagrams. The indices of each network under the four interaction relationships are compared by three models, and the one closest to the middle red line is the optimal model.

图4 影响玉米叶际微生物的显著单核苷酸多样性(SNPs)位点。基于每种关系中每个网络特征值的最优模型进行可视化, 外圆圈的数字代表的是染色体编号, 内层的6个圆圈从里到外依次是连通性(Con)、接近中心度(C(u))、中介中心度(B(u))、偏心率(E(u))、特征向量(G(u))、网页排名(P(u))最优模型的关联结果, 并把显著阈值设为-lg p ≥ 4, 红色线为阈值线, 红色的点为4种关系下6个网络特征值的显著SNPs。

Fig. 4 Significant single-nucleotide polymorphisms (SNPs) locus affecting phyllospheric microbiome of maize. Visualization based on the optimal model of each network eigenvalue in each relationship. The number of the outer circle represents the chromosome number, and the outermost circle represents the association result of the optimal model of connectivity (Con), closeness (C(u)), betweenness (B(u)), eccentricity (E(u)), eigenvector (G(u)), and pagerank (P(u)) from the inside to the outside, the significance threshold is set to -lg p ≥ 4, the red line is the threshold line, and the red point is the significant SNPs locus for the six network feature values under the four relationships.