Fig. 5

Map length ratio between the framework map and the simulated reference map after SeSAM automatic map construction, as a function of the percentage of genotyping errors in a F2 (left panel) or CP (right panel) population of 200 individuals, without and with activating the genotyping error correction option of SeSAM (black circles and red triangles respectively). Data were simulated using the SeSAM function simulatePop() for two chromosomes (100 and 200 cM) with markers regularly spaced at a density of 1 marker/cM, with increasing proportions of genotyping errors uniformly distributed along chromosomes. Lines were obtained by non-linear regression y = a*sqrt(x) + b*x + c (values of (a,b,c) without and with genotyping error correction, leading to respectively (1.09, − 0.06, 0.64) and (0.07, 0.025, 0.89) for F2, and respectively (0.93, − 0.049, 0.64) and (0.20, 0.022, 0.84) for CP). Error bars indicate 95% confidence intervals based on five independent replicates with different seeds for the random number generator used to simulate the data. Dotted lines indicate the theoretical outcome of a perfect genotyping error correction (y = 1)