Figure 1

Performance of the mutation analysis servers grouped by type. Shown is the average percentage of the correctly predicted pathogenic mutants and neutral SNPs. The different servers are divided in groups based on their underlying methods. G = Grantham scores only, MSA = Multiple Sequence Alignment-centred methods (PhD-SNP, Panther, and SIFT), ML = Machine Learning based methods (SNPs&GO, SNAP, MutPred), 3D = structure based methods (nsSNPanalyzer, SNP3D, PolyPhen-2, and HOPE).