Questions and problems?

Nucleic acids, common problems...

Can anyone suggest an easy way to convert protein data base (pdb) files created on amber3a into pdb files readable by xLEaP in the amber4.1 suite of programs. I believe the nucleic acid residue types have been changed in the 1994 force field?

See the pdbconvert program.

A user's listing of common nucleic acid mistakes/factors affecting stability

  1. Usually, the distorted structure produced by not using modified charges for your end base-pair residues. Very common mistakes.
  2. Starting structures are also important if you are running free md.
  3. Use solvated sodium atoms for in vacuo simulation. Modified charges(~0.2) for phosphate backbone produces similar results compared with "fat ions".
  4. Always have it in mind that in vacuo simulation will give you artifacts. Try to run md in water with small ions for 100-200 ps to make sure the final structure is staying there compared to in vacuo simulations.

How to make/use methylphosphonate residues

I would define 2 methylphosphonate residues, e.g. LMP & RMP (one for each phosphate O substitution) using PREP. Then in your link input file (which can be generated by nukit), substitute the desired residue for POM (using a seperate residue file for the modified residue). Let's say you defined the OA-substituted residue as LMP and named the carbon in place of OA, CA (atom name, not type). Take the nucgen pdb file and substitute LMP for POM with the OA->CA atom name change. Now run the pdb and link binary output thru edit using the xray option (see amber4/templates/ and the option to output pdb. EDIT will put the methyl hydrogens in place for you in the output pdb.