서주현 wrote: > I have modeled PLP-dependent aminotransferase. > Making model including PLP was done. But making bond between NZ atom of > lysine and C4A atom of PLP was hard to achieve. How can I make bond > between them?

You are right that creating a true "bond" between a real atom and a block residue atom is tricky, because this requires you to define the topology of the block residue, which is usually not what you want. However, in many cases you don't really need a true bond (which would allow you to build structures from internal coordinates, and is also used for other purposes such as calculating the "distance" between two residues) and can just use a distance restraint instead. For example, to simulate a 1.2A bond between NZ and C4A, you can just use a Gaussian distance restraint. See http://salilab.org/modeller/manual/node26.html and to the special_restraints routine, add

rsr.add(forms.gaussian(group=physical.xy_distance, feature=features.distance(at['NZ:100'], at['C4A:120']), mean=1.2, stdev=0.1))

Obviously you will need to change 100 and 120 for the actual residue numbers of your lysine and block residue (see the .ini file).

Ben Webb, Modeller Caretaker