Fig. 1

Overall genotyping performance of different graph-based tools based on simulated data. The genome graphs of Arabidopsis thaliana (a, b, c), Oryza sativa (d, e, f), Glycine max, (g, h, i), and Zea mays (j, k, l), are constructed based on one reference genome and seven alternative genomes derived by introducing known variants into the reference genome. Paired-end (2 × 150 bp) short reads with 30× depth are simulated for genotyping. For each genotyper, precision is plotted against recall as the genotyping quality threshold varies. Read Depth on variant sites is used as a substitution score when genotype quality is not available. Arrows indicate the circles hidden by other circles in the plot due to identical or nearly identical precision values. Detailed results are also provided in Additional file 2: Table S5