Fig. 6

Example of G23D variant analysis in the catalytic domain of methyl-transferase (DNMT3B). Variants in this protein and specifically in position A585 are known to be related to the ICF syndrome. a DNMT3B has no resolved structures in the PDB, but by using G23D the user can rapidly detect structures of the close homolog DNMT3A (red) and several theoretical models which span the variant (blue). The color gradient indicates the sequence similarity to the PDB protein or to the template protein used for the homology modeling. b Analysis based on the structure of the catalytic domain of DNMT3A which shares 70 % identity with DNMT3B, suggests that the residue in position 585 (shown in black) is completely buried in the protein core and is located on the interface between a helix and a loop. It is positioned close to known pathogenic residues in both ClinVar (red colors) which are automatically being displayed in G23D and HGMD (here manually added in purple). Many other positions appear in Cosmic (yellow). The methyl donor (SAH), whose binding pocket is likely affected by mutations in position 585, appears in magenta. c From the sequence diagram, located in the bottom of the JSmol structure session it can be deduced that the region is very conserved (conservation score of 9) and that it is part of the known Prosite methyltransferase motif (PS51679; in pink). d Contact analysis shows that position 585 has direct contact with some of the known disease related positions shown in panel B, and that the minimal distance of the mutated residue causes steric clashes with two neighbors (marked in arrows). e Stability analysis of the A585T variant (equivalent to position 644 in PDB 4u7p ATOM section) using I-Mutant-2, directly accessed from G23D, suggests decreased stability of the mutant structure