Tim wrote:
> my apologies for jumping in late to this discussion.
>
> the offset scheme in the RasMol doc appears to do a good partial job -
> showing helix and turn vs all other hbonds.

Agreed.

> my fear is that many of the
> 'biologically significant' hbonds (i.e. those with a functional role,
> not just structural scaffolding) will just be yellow, with no way to
> tease them apart.  this includes not only beta sheet hbonds, hbonds
> within a chain, hbonds between chains, hbonds between protein and
> non=protein (including solvent), etc.

Yes.

The only *good* news (if you can call it that) is that most of these
hbonds don't actually exist. All we calculate today are backbone hbonds
within the same chain.

> also, what happens (I haven't tested) if there is an hbond between
> residue 5A and 8B?

Not sure exactly what you mean here.

If you are concerned that insertion codes might mess things up, then I
don't think you need to worry ... I believe that Jmol will do the right
thing. The groupID is not used for this calculation.

> it would be colored magenta, if I'm not mistaken,
> but that gives the incorrect impression that these are close to each
> other in the primary sequence.  and what happens when residue numbers
> are not continuous?

Residue numbers are not used at all.

> or when hbonds are between protein and
> non-proteins?  the last case seems to me to be very important.

Modified amino acids in the backbone should work OK.

Remember, I am only talking about backbone hbonds at this point.


> I wonder if building boolean selections is not a more useful way to
> address the different types of hydrogen bonds.  we did something like
> this with the Jmol 'within' command, which now lets you expand a
> selection to include the groups that contain the selected atoms.
>
> can one identify (iow, select) the partners in a hydrogen bond
> interaction?  for example, 'select donor' or 'select acceptor'?
> preferably at the atomic level, since 'within' can expand out the entire
> group.  then we could select not only the bonds, but the bonding
> partners as well, based on type.

Let's talk about this later.

> Miguel, how muddy are the waters now?

Somewhat murky, but not too bad :-)


I am going to try to restate the question that I would like to have an
answer to.

- Like RasMol, Jmol only supports backbone hbonds in proteins. On the
nucleotide side, it only supports those *classical* nucleotide hbonds that
hold the helix together, whatever they are called.

- In RasMol, the script command 'color hbonds type' will make random coil
hbonds, beta-sheet hbonds, and these nucleotide hbonds all the same color
... yellow.

Q: Is that the desired behavior?


Miguel



-------------------------------------------------------
This SF.Net email is sponsored by: SourceForge.net Broadband
Sign-up now for SourceForge Broadband and get the fastest
6.0/768 connection for only $19.95/mo for the first 3 months!
http://ads.osdn.com/?ad_id=2562&alloc_id=6184&op=click
_______________________________________________
Jmol-developers mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jmol-developers

Reply via email to