Fig. 6: Relationships between MIcomplementarity and estimated genome size and biological function proportion.
a Correlation (Spearman’s rank correlation test, two-sided) between normalized average abundance of each temperature group and the estimated genome size. b Correlation matrix on proportions of twelve COG biological functions in MAGs. The lines between estimated genome size and each COG biological function indicated Spearman’s correlation (two-sided). The color shade of the lines represented the strength of the correlation, and the thickness represented the confidence level. c Effect of COG biological function proportion on MIcomplementarity by linear mixed-effects models (LMMs) fit by restricted maximum likelihood (REML) estimation. Data are presented as mean values ± standard errors of the estimated effect sizes. Bar length and error bar indicated the mean values and standard errors of the estimated effect sizes. Significance test was conducted using a t-test of Satterthwaite’s methods (number of observations n = 80,200 for non-self-loop pairwise interactions between 401 MAGs). Significant effects are represented by asterisks: ***p < 0.001, **p < 0.01, ‘ns’ stands for not significant. d–f Correlation between MIcomplementarity and genome size difference. The fitted curve was modeled using the generalized linear model. The correlation was tested using Spearman’s rank correlation test (two-sided).
