[Chimera-users] Re: bug? (HETATMs)

[Eric Pettersen]

On Saturday, October 4, 2003, at 10:21 AM, solomon wrote:

> when i open a pdb format file like below:
> HETATM .................................
> HETATM .................................
> HETATM .................................
> ...
> ...
> ...
> HETATM .................................
>  
> then I find that all the HETATM type atoms are bonding each other. All 
> the HETATM need a ter card to split them. As for HETATM type, I think 
> maybe there is some wrong with chimera.
>  best regards
> gaokeng xiao

Dear Gaokeng,
	As far as know, Chimera will connect HETATM segments of all standard 
PDB files (i.e. those distributed by the RCSB) correctly.  That being 
said, Chimera has to do a lot of guessing when it comes to connecting 
non-standard PDB files (which I'm pretty sure is the kind of file 
you're talking about here).
	One thing to keep in mind with HETATMs is that they don't necessarily 
imply a chain break.  Also, they are supposed to have CONECT records 
detailing their connectivity.  So for instance, part of PDB entry 1f1j 
looks like this:

ATOM   3688  OG  SER B 602      58.714  25.290  29.976  1.00 66.81      
      O
TER    3689      SER B 602
HETATM 3690  C   ACE C 701      29.751  29.860  -0.650  1.00 52.20      
      C
HETATM 3691  O   ACE C 701      28.552  30.100  -0.812  1.00 55.72      
      O
HETATM 3692  CH3 ACE C 701      30.801  30.183  -1.675  1.00 52.54      
      C
ATOM   3693  N   ASP C 702      30.252  29.308   0.427  1.00 51.48      
      N

	The ACE residue in the HETATM records is connected to residue 702 (as 
detailed in the CONECT records for the file).  So in summary, you can 
have sequences like these where the HETATM residues are connected:

ATOM
ATOM
HETATM
HETATM
HETATM
ATOM
ATOM

-or-

HETATM
HETATM
HETATM
ATOM
ATOM

	Now the really big problem is that the SCOP web site returns "PDB" 
files with just these kinds of sequences but with _no_ CONECT records!  
So, Chimera guesses as best it can in this situation.  The rule of 
thumb that it uses is that HETATMs are connected unless they occur 
somewhere after a TER card with no intervening ATOM record.  Explicit 
CONECT records can add connectivity.
	So for your file of all HETATMs where nothing should be connected, 
there is a silver lining:  in order to get nothing connected you just 
have to add a _single_ TER card at the beginning of the file, like this:

TER
HETATM .................................
HETATM .................................
HETATM .................................

--Eric