Fig. 2

Different microbial communities are simulated to test the performance of the proposed method. (a) Results for microbial communities with 10 species. The three histograms on the left show the abundance distributions of different simulated communities. The middle panel shows the estimation results of distinct k-mer count. Each bar represents an estimation result based on a synthetic metagenomic sample and the error bar shows the 95% bootstrap confidence interval of the estimation. The black dash line is the true distinct k-mer count. The right panel shows how the relative error goes as the initial coverage increases (k = 20). (b) The same as (a) except that the species number is 50. (Note that some of the samples with 10 species are not shown in the barplot, see Additional file 1: Figure S1 for all samples with 10 species)