Comparison between LAP scores and the rankings of the top assemblies generated in the Assemblathon 1 competition. The colors represent the relative ranking provided by the individual metrics (best - green, worst - red): log average probability (LAP), overall coverage (Cov tot), contig path NG50 (CPNG50), sum of all rankings from Assemblathon 1 (Overall), weighted median contig size based on estimated genome size (NG50), coverage within coding sequences (Cov genic), scaffold path NG50 (SPNG50), length for which half of any two valid columns in the assembly are correct in order and orientation (CC50), weighted median contig size based on total assembly size (N50), proportion of columns with a copy number error (Copy num), total substitution errors per correct bit (Subs), and sum of structural errors (Struct). Column descriptions and underlying data obtained from Table 3 in Earl et al. . Columns are sorted according to the level of concordance with the LAP ranking. De novo measures are highlighted in bold.