[MMTK] PDB files with modified residues like MSE (selenomethionine)

[Peter]

Hello again,

I'm trying to load PDB file 1PYM, but have run into a problem with 
modified amino acid residues, held as HETATM entries.

MMTK seems to break the peptide chain at these points (e.g. Mse14), 
giving a chain covering "Chain A" from residue Val5 to Gln13, and a new 
chain starting Leu15 to Ile23, etc

sample code:

import MMTK.Proteins
protein = MMTK.Proteins.Protein("1PYM.pdb", model="no_hydrogens")
assert len(protein.chains)==16

except from 1PYM file, Mse14 on chain A:-

ATOM     74  NE2 GLN A  13      20.376  71.210  49.248  ..
HETATM   75  N   MSE A  14      20.751  70.044  55.325  ...
HETATM   76  CA  MSE A  14      20.869  68.953  56.262  ...
HETATM   77  C   MSE A  14      19.901  69.044  57.444  ...
HETATM   78  O   MSE A  14      19.240  68.075  57.773  ...
HETATM   79  CB  MSE A  14      22.285  68.827  56.761  ...
HETATM   80  CG  MSE A  14      23.226  68.367  55.726  ...
HETATM   81 SE   MSE A  14      25.099  68.515  56.262  ...
HETATM   82  CE  MSE A  14      25.283  66.996  57.406  ...
ATOM     83  N   LEU A  15      19.811  70.203  58.091  ...

The use of HETATM rather than ATOM means this is a modified residue - In 
this case, MSE (selenomethionine) is a modified MET (methionine) where 
there is a selenium in place of sulfur.  (See also the HET and HETNAM 
entries in the file header).

I understand that the AMBER forcefields don't cover these cases, but in 
order to get MMTK to load the file "properly" without breaking the 
peptides do I just need to define a new selenomethionine group in the 
database?

Or, due to the use of HETATM rather than ATOM, is this more complicated?

Thank you

Peter