Miguel wrote:

Jan wrote:



Are you just talking about the AIB groups 22 & 23 ?

Do you think that we should allow O1 as an acceptable atom
name for the carbonyl oxygen?



no, I suggest the combination of element type and connection pattern to
determine an amino acid



The connection pattern is currently used to identify an AminoMonomer.

For a _group_ to be an AminoMonomer it must contain the following:
* atoms with the following names:
   - N
   - CA
   - C
   - O
* the atoms must also be connected correctly
   - N + CA
   - CA + C
   - C + O

Note that I cannot check bond orders for these because .pdb files do not
contain bond order information.



0. Ala Arg Asp ...



I think that you are saying that the *group name* should be used to determine whether or not something is an amino acid.

At this point I think that would be very difficult. There is too much code
that depends upon the presence of all the key atoms. In particular, all of
the secondary structure shapes are dependant upon the presence of CA-C=O


it is OK to have the amino set



1. *.CA only: within(4.2, *.CA) != (*.N, *.O, *.OXT, *.C)



I do not understand what you are suggesting here.

The current support for AlphaMonomer (for alpha-carbon-only models) is as
follows:

* a _single_ atom in a group whose name is CA


you should add that *.CA must be carbon e.g.
http://www.imb-jena.de/cgi-bin/SCOPlnk.exe?JMOL=1a0j
select protein and [CA]
selects the Ca atom



2. default naming *.CA *.C *.N *.O



These names are currently required.



3. connection C?-[-N?-C?-C?=(-O?] -N?) only the CAlpha should branch
within the same residue (and N ? case of Pro)



I do not understand what you are saying here.


I tried to describe what makes an amino acid only with elements and bonding pattern



If we make this change, do you think that it will cause
other problems?




if we include the connection that made an amino acid we should get amino
acids?



As described above, connectivity is currently used to determine if a group is an AminoMonomer.



But there are ~1000 HETERO components with N?-C?-C?-O? connection
(without checking branching e.g. Guanosine derivate: OMG N7
C5 C6 O6)



I do not understand what you are saying here.


I am contradictory :-) trying to *start discussion*, whether element bonding and branching is enough to define, what is an amino acid, and should be implemented instead of a naming dependent pattern?
Regards, Jan



------------------------------------------------------- This SF.Net email is sponsored by BEA Weblogic Workshop FREE Java Enterprise J2EE developer tools! Get your free copy of BEA WebLogic Workshop 8.1 today. http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click _______________________________________________ Jmol-developers mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jmol-developers

Reply via email to