Re: [Jmol-users] Protein Hydrogen Bonds again
Bob wrote: ah -- ok, I see. Well, that sounds terrific. I'd be interested just in peeking at the C code, then, in that case. Just to see what sort of considerations you made. Who knows. It might be an almost trivial addition to Jmol. I'll send it you. David ___ Dr. David P. Leader, Faculty of Biomedical Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK Phone: +44 (0)141 330 5905 http://doolittle.ibls.gla.ac.uk/leader The University of Glasgow, charity number SC004401 ___ - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100url=/ ___ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Bob wrote: It might be interesting to see how the HBplus program works. I see http://www.csb.yale.edu/userguides/datamanip/hbplus/algorithm.txt but that suggests there is no energy calculation here, just a set of angle/distance rules. HOWEVER -- NOTE!!! The HBPlus program itself is not public and its code should NOT be used in any way in any relation to Jmol. See http://www.bio.net/bionet/mm/bio-soft/1994-August/008380.html -- unless the authors have relaxed those publication parameters. Yes, I know. We used that under license (!) to populate our db and search for motifs. However, the C program I mentioned is our property, does energy calculations and would seem as good as HB-plus. We would make any port to java open source so there would be no problem. I'm going to write a proposal tomorrow. Whether we get an interested student is another matter. David ___ Dr. David P. Leader, Faculty of Biomedical Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK Phone: +44 (0)141 330 5905 http://doolittle.ibls.gla.ac.uk/leader http://motif.gla.ac.uk/ The University of Glasgow, charity number SC004401 ___ - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100url=/___ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
ah -- ok, I see. Well, that sounds terrific. I'd be interested just in peeking at the C code, then, in that case. Just to see what sort of considerations you made. Who knows. It might be an almost trivial addition to Jmol. Bob On Mon, Nov 10, 2008 at 11:36 AM, David Leader [EMAIL PROTECTED]wrote: Bob wrote: It might be interesting to see how the HBplus program works. I see http://www.csb.yale.edu/userguides/datamanip/hbplus/algorithm.txt but that suggests there is no energy calculation here, just a set of angle/distance rules. HOWEVER -- NOTE!!! The HBPlus program itself is not public and its code should NOT be used in any way in any relation to Jmol. See http://www.bio.net/bionet/mm/bio-soft/1994-August/008380.html -- unless the authors have relaxed those publication parameters. Yes, I know. We used that under license (!) to populate our db and search for motifs. However, the C program I mentioned is our property, does energy calculations and would seem as good as HB-plus. We would make any port to java open source so there would be no problem. I'm going to write a proposal tomorrow. Whether we get an interested student is another matter. David ___ Dr. David P. Leader, Faculty of Biomedical Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK Phone: +44 (0)141 330 5905 http://doolittle.ibls.gla.ac.uk/leader http://motif.gla.ac.uk/ The University of Glasgow, charity number SC004401 ___ - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100url=/ ___ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users -- Robert M. Hanson Professor of Chemistry St. Olaf College 1520 St. Olaf Ave. Northfield, MN 55057 http://www.stolaf.edu/people/hansonr phone: 507-786-3107 If nature does not answer first what we want, it is better to take what answer we get. -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900 - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100url=/___ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
right -- as I said in an earlier message, I used 2.5. Try this:
connect 2.5 {_H and connected(_O or _N)} {(_O or _N) and not within(group,
_1)} HBONDS CREATE
On Sun, Nov 9, 2008 at 2:54 PM, Angel Herráez [EMAIL PROTECTED] wrote:
David, have you measured the length of the Hbonds you want to show up?
You might nedd some longer spec.
I've just tested
connect 1.8 {_H and connected(_O or _N)} {_O or _N} HBONDS CREATE;
with
http://doolittle.ibls.gla.ac.uk/top500H/1osaH.pdb
and Jmol 11.6.3 and it creates and draws 12 new Hbonds
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Angel wrote:
David, have you measured the length of the Hbonds you want to show up?
You might nedd some longer spec.
I've just tested
connect 1.8 {_H and connected(_O or _N)} {_O or _N} HBONDS CREATE;
with
http://doolittle.ibls.gla.ac.uk/top500H/1osaH.pdb
and Jmol 11.6.3 and it creates and draws 12 new Hbonds
Thanks Angel. I've tried that, but the 12 don't include the important
ones for me.
The point is that these are Hbond identified on the basis of energy
using the HBplus program, not just distance, and deposited in my
database. The motifs I visualize are identified by angle criteria
together with the existance of these HBonds, again from my database.
Bob wrote:
connect 2.5 {_H and connected(_O or _N)} {(_O or _N) and not within
(group,
_1)} HBONDS CREATE
OK, that probably includes the ones I want (difficult to see because
the hydrogens are not there) but the number of background hydrogen
bonds make it look awful.
Sorry to have wasted your time, but I'm now convinced that the SQL
route is the way I'm going to have to go. I've tested a query that
finds the Hbonds associated with the atoms in a three-residue motif
which takes 4 secs to run (wasn't so difficult to formulate after all
- only involved three tables). There are about 8 hbonds in the
resulting table, two of which have the same id, and I can draw using
their atom numbers (which are retrieved). If I can do some DB
indexing to reduce the query time by an order of magnitude I think
I'll be able to produce what I want with another query on the
resulting table.
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
David, have you measured the length of the Hbonds you want to show up?
You might nedd some longer spec.
I've just tested
connect 1.8 {_H and connected(_O or _N)} {_O or _N} HBONDS CREATE;
with
http://doolittle.ibls.gla.ac.uk/top500H/1osaH.pdb
and Jmol 11.6.3 and it creates and draws 12 new Hbonds
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Bob wrote: Sure, but -- you said you had the H atoms Are you saying you might want to move them? On Thu, Nov 6, 2008 at 11:01 AM, David Leader [EMAIL PROTECTED]wrote: However I still might see if there is a student up to integrating the C code. My colleague tells me that this is quite sophisticated. For example it calculates the H atoms and deals with the uncertainty in serines and threonines, by considering three possible rotamers. David Sure, for my immediate purposes with the proteins I have the connect script might be ok (haven't got it working yet, but I'm busy teaching and watching football at the moment), but in the long term for general use a function that calculates the protein H-bonds properly (ie calculates energies, as our program does) would be of general benefit. Any student working on it wouldn't start properly until next May, so it's longer term. However I need to suggest a project by the end of next week - hence my query. I certainly will look at the connect script over the weekend and get back to you if necessary. David ___ Dr. David P. Leader, Faculty of Biomedical Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK Phone: +44 (0)141 330 5905 http://doolittle.ibls.gla.ac.uk/leader http://motif.gla.ac.uk/ The University of Glasgow, charity number SC004401 ___ - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100url=/___ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Bob,
Sorry but I can't get the script line you sent to work.
To check it wasn't my version of Jmol I went to your tutorial site
and got the example page for connect. Then I:
1. Uploaded an example file with Hs: http://doolittle.ibls.gla.ac.uk/
top500H/1osaH.pdb
2. Pasted some script in the console to get the appearance I use:
set perspectiveModel 11; set background white; cpk off; restrict
protein; wireframe off; backbone 0.3; set hydrogen off; centre
selected; zoom (selected) 0;
That worked fine when I hit execute.
3. Then pasted into the console your expression:
connect 1.8 {H and connected(_O or _N)} {_O or _N} HBONDS CREATE;
On execute I got,
0 new bonds; 0 modified
Script completed
Jmol script terminated successfully
Jmol script terminated
(Same with hbonds on after this.)
Same if I left the appearance as it comes up initially.
As you can see if you download it, the file definitely has Hs added
(with the Richardson lab 'reduce' program).
Any idea what I am doing wrong?
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
http://motif.gla.ac.uk/
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
David,
connect 1.8 {H andetc.
is not correct; you need _H or hydrogen, not H
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Angel wrote:
David,
connect 1.8 {H andetc.
is not correct; you need _H or hydrogen, not H
So I do. It works now. Many thanks!
I'm very curious to see whether it does the right thing with the
relevant motifs. But it's been a hard day in the bioinformatics lab,
so it will have to wait until after supper and probably another day.
I'll report back.
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
http://motif.gla.ac.uk/
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
OOPS! Thanks, Angel -- right, _H, not H. I was working too fast.
Bob
On Fri, Nov 7, 2008 at 11:27 AM, David Leader [EMAIL PROTECTED]wrote:
Angel wrote:
David,
connect 1.8 {H andetc.
is not correct; you need _H or hydrogen, not H
So I do. It works now. Many thanks!
I'm very curious to see whether it does the right thing with the relevant
motifs. But it's been a hard day in the bioinformatics lab, so it will have
to wait until after supper and probably another day. I'll report back.
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
http://motif.gla.ac.uk/
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
with that sample file, this looks better to me:
connect 2.5 {_H and connected(_O or _N)} {(_O or _N) and not within(group,
_1)} HBONDS CREATE
That last bit guarantees that the bonds are inter-residue bonds, not
intra-residue. It says, and not within the group of the atom selected
already
Bob
On Fri, Nov 7, 2008 at 11:50 AM, Robert Hanson [EMAIL PROTECTED] wrote:
OOPS! Thanks, Angel -- right, _H, not H. I was working too fast.
Bob
On Fri, Nov 7, 2008 at 11:27 AM, David Leader [EMAIL PROTECTED]wrote:
Angel wrote:
David,
connect 1.8 {H andetc.
is not correct; you need _H or hydrogen, not H
So I do. It works now. Many thanks!
I'm very curious to see whether it does the right thing with the relevant
motifs. But it's been a hard day in the bioinformatics lab, so it will have
to wait until after supper and probably another day. I'll report back.
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
http://motif.gla.ac.uk/
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the
world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
OK. Keep me informed. The energy calculation that Jmol uses in general for
hydrogen bond calculation is specific to carbonylCO-amideNH interactions and
looks like the following. I'm sorry I don't have a literature reference to
this. If you know it, please let me know -- I'd like to add that as a
comment to the code. I remember some time back having to modify it, though,
because it had some failures. I can't remembe at this moment exactly what I
had to change. So it's not exactly what is published, I think.
The formula is:
QConst = -332 * 0.42 * 0.2 * 1000
energy = (QConst / distOH - QConst / distCH)
+ (QConst / distCN - QConst / distON)
isHbond = (distCN2 distCH2 distOH = 3.0f energy = -500)
which is basically just a little formal charge electrostatic
repulsion/attraction approximation putting +1 charge at C and H and -1
charge at O and N.
Bob
from
http://jmol.svn.sourceforge.net/viewvc/jmol/trunk/Jmol/src/org/jmol/modelsetbio/AminoPolymer.java?view=markup
private final static double QConst = -332 * 0.42 * 0.2 * 1000;
private int calcHbondEnergy(Atom nitrogen, Point3f nitrogenPoint,
Point3f hydrogenPoint, AminoMonomer target) {
Point3f targetOxygenPoint = target.getCarbonylOxygenAtomPoint();
float distON2 = targetOxygenPoint.distanceSquared(nitrogenPoint);
if (distON2 minimumHbondDistance2)
return 0;
float distOH2 = targetOxygenPoint.distanceSquared(hydrogenPoint);
if (distOH2 minimumHbondDistance2)
return 0;
Point3f targetCarbonPoint = target.getCarbonylCarbonAtomPoint();
float distCH2 = targetCarbonPoint.distanceSquared(hydrogenPoint);
if (distCH2 minimumHbondDistance2)
return 0;
float distCN2 = targetCarbonPoint.distanceSquared(nitrogenPoint);
if (distCN2 minimumHbondDistance2)
return 0;
double distOH = Math.sqrt(distOH2);
double distCH = Math.sqrt(distCH2);
double distCN = Math.sqrt(distCN2);
double distON = Math.sqrt(distON2);
int energy = (int) ((QConst / distOH - QConst / distCH + QConst / distCN
- QConst
/ distON));
boolean isHbond = (distCN2 distCH2 distOH = 3.0f energy =
-500);
return (!isHbond ? 0 : energy -9900 ? -9900 : energy);
}
Jmol creates up to two hydrogen bonds to a
On Fri, Nov 7, 2008 at 3:16 AM, David Leader [EMAIL PROTECTED] wrote:
Bob wrote:
Sure, but -- you said you had the H atoms Are you saying you might want
to move them?
On Thu, Nov 6, 2008 at 11:01 AM, David Leader [EMAIL PROTECTED]
wrote:
However I still might see if there is a student up to integrating the C
code. My colleague tells me that this is quite sophisticated. For example
it
calculates the H atoms and deals with the uncertainty in serines and
threonines, by considering three possible rotamers.
David
Sure, for my immediate purposes with the proteins I have the connect script
might be ok (haven't got it working yet, but I'm busy teaching and watching
football at the moment), but in the long term for general use a function
that calculates the protein H-bonds properly (ie calculates energies, as our
program does) would be of general benefit. Any student working on it
wouldn't start properly until next May, so it's longer term. However I need
to suggest a project by the end of next week - hence my query.
I certainly will look at the connect script over the weekend and get back
to you if necessary.
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
http://motif.gla.ac.uk/
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
Re: [Jmol-users] Protein Hydrogen Bonds again
Bob wrote:
David -- none of the work I just did will be of help to you, since
you have
the H atoms already present. But I'm glad you started this discussion
anyway, because that will certainly be of interest to others.
I really think a simple command such as
connect 1.8 {H and connected(_O or _N)} {_O or _N} HBONDS CREATE
is all you need.
Bob
Ok, I'll try that and let you know how I get on.
However I still might see if there is a student up to integrating the
C code. My colleague tells me that this is quite sophisticated. For
example it calculates the H atoms and deals with the uncertainty in
serines and threonines, by considering three possible rotamers.
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
http://motif.gla.ac.uk/
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Sure, but -- you said you had the H atoms Are you saying you might want
to move them?
On Thu, Nov 6, 2008 at 11:01 AM, David Leader [EMAIL PROTECTED]wrote:
Bob wrote:
David -- none of the work I just did will be of help to you, since you have
the H atoms already present. But I'm glad you started this discussion
anyway, because that will certainly be of interest to others.
I really think a simple command such as
connect 1.8 {H and connected(_O or _N)} {_O or _N} HBONDS CREATE
is all you need.
Bob
Ok, I'll try that and let you know how I get on.
However I still might see if there is a student up to integrating the C
code. My colleague tells me that this is quite sophisticated. For example it
calculates the H atoms and deals with the uncertainty in serines and
threonines, by considering three possible rotamers.
David
___
Dr. David P. Leader, Faculty of Biomedical Life Sciences,
University of Glasgow, Glasgow G12 8QQ, UK
Phone: +44 (0)141 330 5905
http://doolittle.ibls.gla.ac.uk/leader
http://motif.gla.ac.uk/
The University of Glasgow, charity number SC004401
___
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
On Thu, November 6, 2008 12:01 pm, David Leader wrote: However I still might see if there is a student up to integrating the C code. My colleague tells me that this is quite sophisticated. For example it calculates the H atoms and deals with the uncertainty in serines and threonines, by considering three possible rotamers. Calculation of reasonable H positions is indeed non-trivial. Experimental observation (using X-rays) of electron density associated with H atom positions is not commonly possible for protein structures; the necessary resolution (better than 0.9A) simply isn't available for most crystals. Even in higher resolution small-molecule work the actual positions on N and O atoms are as much inferred as observed. So, in most cases, some considerable effort both computationally and analytically is needed to produce a self-consistent and chemically reasonable H-network; no simple two or three-body geometry calculation will produce the correct results all or even most of the time. I suppose there might be some value in having Jmol calculate display a *potential* H-bonding scheme as long as it is understood to be a cartoon only. Rich - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100url=/ ___ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
oops -- that didn't work -- will have this corrected later this morning.
Looks like 90 degrees is more reasonable -- 120 cuts out way too many.
On Tue, Nov 4, 2008 at 8:25 PM, Robert Hanson [EMAIL PROTECTED] wrote:
should be adjustable -- but I haven't actually tested it. Just thought I
would get that out there for inspection.
David -- to a first approximation, of course, if you have the H atoms and
you can live with just distance and a parameter, it's just a connect command
to produce all those hydrogen bonds. If you need more, then it's just a
matter of defining what angle requirements one might want.
Bob
On Tue, Nov 4, 2008 at 7:49 PM, Thomas Stout [EMAIL PROTECTED]wrote:
It looks nice! -- is the angle adjustable? I tried setting it to other
values (even extreme values) and didn't see any change in the created H
bonds
-Tom
On Tue, Nov 4, 2008 at 4:40 PM, Robert Hanson [EMAIL PROTECTED] wrote:
see
http://chemapps.stolaf.edu/jmol/docs/examples-11/new.htm
The min/max distances are same as for other H bonds:
hbondMax = 3.25f;
hbondMin = 2.5f;
These values aren't currently adjustable.
comments requested
On Tue, Nov 4, 2008 at 5:23 PM, Robert Hanson [EMAIL PROTECTED]wrote:
Possibly. Is that documented or discussed somewhere in the literature?
I could probably implement that tonight while watching the election
returns then let you try it out. I propose:
calculate hbonds {atomset1} {atomset2}
maybe with some settings for minimum and maximum distances.
Bob
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the
world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the
world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
OK, that's uploaded. See
http://chemapps.stolaf.edu/jmol/docs/examples-11/new.htm and
http://chemapps.stolaf.edu/jmol/docs/examples-11/Jmol-11.zip
On Wed, Nov 5, 2008 at 8:59 AM, Robert Hanson [EMAIL PROTECTED] wrote:
oops -- that didn't work -- will have this corrected later this morning.
Looks like 90 degrees is more reasonable -- 120 cuts out way too many.
On Tue, Nov 4, 2008 at 8:25 PM, Robert Hanson [EMAIL PROTECTED] wrote:
should be adjustable -- but I haven't actually tested it. Just thought I
would get that out there for inspection.
David -- to a first approximation, of course, if you have the H atoms and
you can live with just distance and a parameter, it's just a connect command
to produce all those hydrogen bonds. If you need more, then it's just a
matter of defining what angle requirements one might want.
Bob
On Tue, Nov 4, 2008 at 7:49 PM, Thomas Stout [EMAIL PROTECTED]wrote:
It looks nice! -- is the angle adjustable? I tried setting it to other
values (even extreme values) and didn't see any change in the created H
bonds
-Tom
On Tue, Nov 4, 2008 at 4:40 PM, Robert Hanson [EMAIL PROTECTED]wrote:
see
http://chemapps.stolaf.edu/jmol/docs/examples-11/new.htm
The min/max distances are same as for other H bonds:
hbondMax = 3.25f;
hbondMin = 2.5f;
These values aren't currently adjustable.
comments requested
On Tue, Nov 4, 2008 at 5:23 PM, Robert Hanson [EMAIL PROTECTED]wrote:
Possibly. Is that documented or discussed somewhere in the literature?
I could probably implement that tonight while watching the election
returns then let you try it out. I propose:
calculate hbonds {atomset1} {atomset2}
maybe with some settings for minimum and maximum distances.
Bob
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the
world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the
world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
That's very nice -- I'm quite pleased with that feature
Thanks Bob!
-Tom
On Wed, Nov 5, 2008 at 10:56 AM, Robert Hanson [EMAIL PROTECTED] wrote:
OK, that's uploaded. See
http://chemapps.stolaf.edu/jmol/docs/examples-11/new.htm and
http://chemapps.stolaf.edu/jmol/docs/examples-11/Jmol-11.zip
On Wed, Nov 5, 2008 at 8:59 AM, Robert Hanson [EMAIL PROTECTED] wrote:
oops -- that didn't work -- will have this corrected later this morning.
Looks like 90 degrees is more reasonable -- 120 cuts out way too many.
On Tue, Nov 4, 2008 at 8:25 PM, Robert Hanson [EMAIL PROTECTED] wrote:
should be adjustable -- but I haven't actually tested it. Just thought I
would get that out there for inspection.
David -- to a first approximation, of course, if you have the H atoms and
you can live with just distance and a parameter, it's just a connect command
to produce all those hydrogen bonds. If you need more, then it's just a
matter of defining what angle requirements one might want.
Bob
On Tue, Nov 4, 2008 at 7:49 PM, Thomas Stout [EMAIL PROTECTED]wrote:
It looks nice! -- is the angle adjustable? I tried setting it to other
values (even extreme values) and didn't see any change in the created H
bonds
-Tom
On Tue, Nov 4, 2008 at 4:40 PM, Robert Hanson [EMAIL PROTECTED]wrote:
see
http://chemapps.stolaf.edu/jmol/docs/examples-11/new.htm
The min/max distances are same as for other H bonds:
hbondMax = 3.25f;
hbondMin = 2.5f;
These values aren't currently adjustable.
comments requested
On Tue, Nov 4, 2008 at 5:23 PM, Robert Hanson [EMAIL PROTECTED]wrote:
Possibly. Is that documented or discussed somewhere in the literature?
I could probably implement that tonight while watching the election
returns then let you try it out. I propose:
calculate hbonds {atomset1} {atomset2}
maybe with some settings for minimum and maximum distances.
Bob
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the
world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the
world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
Re: [Jmol-users] Protein Hydrogen Bonds again
David -- none of the work I just did will be of help to you, since you have
the H atoms already present. But I'm glad you started this discussion
anyway, because that will certainly be of interest to others.
I really think a simple command such as
connect 1.8 {H and connected(_O or _N)} {_O or _N} HBONDS CREATE
is all you need.
Bob
On Tue, Nov 4, 2008 at 6:06 PM, David Leader [EMAIL PROTECTED] wrote:
Thanks for the response to my message. I've been tied up with
something and now it's quite late in Europe so I'll just send a brief
reply to a couple of points and any further info later.
1. My interest is only in proteins where the hydrogens have been
added. The database of the proteins I wish to display is
predominantly the Richardson 500, where the resolution is never more
than 2 angstroms, all the hydrogens have been added, and ambiguous
residues checked manually. We are only interested in displaying H-
bonds we are pretty sure about.
The suggestion of writing extra scripts seems unattractive at first
sight. At the moment, after a search I am allowing a user to fire up
Jmol showing the motifs in the context of the protein all generated
on the fly. It's tricky enough to generate javascript controls for a
particular situation in the context of the pdb file and the residues
in the motifs. I haven't done anything with Jmol for two years (my
web app uses 10.2) so perhaps I am overestimating the difficulty of
the problem.
David
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
I am not knowledgeable about the algorithms that are out there (perhaps someone like Warren DeLano and others might chime in), but I'd like to throw in my 2 cents (currently approaching the true value of my opinion with the state of the US dollar!). I'd rather see Jmol do *something *in this regard rather than nothing (or nearly nothing), which is the current offering (limited to mainchain interactions of pre-defined secondary structure). To me, a 90+ percent solution is preferable than the 100 percent solution (my opinion; I know others will disagree). It could be couched in terms of possible hydrogen bonds or geometric interactions or something less committal than claiming accurate H-bonds, but at least if there is a set of dotted lines between heteroatoms that may interact via a localized hydrogen, then people can more easily begin to ask questions. Swiss PDB Viewer (aka, DeepView) does a nice job in this regard. While there are a multitude of possibly correct locations for hydrogens in protein structures, the geometry constrains the number of possibilities for true hydrogen mediated electrostatic interactions. Once upon a time I wrote some crude scripting for PyMOL that created eactly what is being asked for here. The basic premise was: if two appropriate heteroatoms are spatially located within a set range of separations (1.8-3.4A) and disposed to one another within a certain angular spread (120-180 degrees) relative to the atoms to which they are covalently bonded, then they are likely candidates for a hydrogen bonded interaction. This gets you to within a few percent error and it gives the viewer something to consider. (I made no attempt to address extended interactions and flipping of sidechains, etc and I believe that that is not the role of Jmol either) Food for thought(low calorie though it may be!):) -Tom On Tue, Nov 4, 2008 at 10:57 AM, Robert Hanson [EMAIL PROTECTED] wrote: On Tue, Nov 4, 2008 at 9:36 AM, David Leader [EMAIL PROTECTED]wrote: Question 1: Jmol will display main-chain main-chain H-bonds in proteins read from pdb files in which this information (or implied info as secondary structure) does not appear to be present. Am I right in thinking that Jmol calculates these, presumably while it is calculating secondary structure? Secondary structure is determined by phi and psi angles, not hydrogen bonding. Question 2: It was suggested to me that as I have the side-chain H-bond info in a relational database (from the HBplus program) I could add this to a pdb file, specifying CONECTS, and Jmol would interpret this. However, it seems it would interpret this as covalent bonding rather than H-bonding. What I want is to be able to turn this on and off in the same way that I can script the current bonds. Anyone shed some light on this? This is an excellent idea. Just use connect ... ... HBOND. If you can have the program output Jmol connect commands, that would be ideal. Otherwise, using a Jmol script perhaps you could read the file into a Jmol variable, parse the data, and create its own connect commands. Question 3: If the route suggested in 2 is not really an option, what is the team's attitude to adding code to Jmol. A colleague has just found some C code a student wrote many years ago that calculates side-chain H-bonds. The advantage of this is that we have the source and porting from C to Java would be less difficult than porting from some other languages (e.g. Fortran). If I proposed it as a 15-week project for a Masters student in Computing Science or IT, would this seem feasible and would it be welcomed by the team? I think 2 is a very very good option. If we wanted to do something more in the area of calculation, it's important to explore how other programs handle this. If there is an accepted, published method, it would be preferable to implement that rather than code from a student many years ago. That's because then we can state the Jmol implements recognizable Algorithm X. In any case, the method needs to be easily described and implemented. Jmol has powerful tools that can be applied, so it's a bit more complicated than just replicating some C code. One of the issues is that most PDB files don't have the actual hydrogen atoms. My initial attempts at this convinced me it is a rather difficult problem when the H atoms are not there. So think about whether you want to limit it to cases where the H atoms are present or not. Hydroxyl groups and histidines come to mind; pH and actual protonation state, etc. Many of these issues are dealt with far better by programs such as those in the AMBER tools package, so another possibility would be to leave that problem to others and only focus on the visualization of already-determined hydrogen bonds. Bob David ___ Dr. David P. Leader, Faculty of Biomedical Life Sciences, University of Glasgow,
Re: [Jmol-users] Protein Hydrogen Bonds again
For the angles, let me try to describe what I mean:
you have two hetero atoms, X and Y, located less than 3.4A from each other.
one of those atoms bears a hydrogen that is not present in the coordinate
file, let's not worry about which one.
each of X and Y are further covalently bonded to at least one other protein
atom, X2 and/or Y2.
The angle created by X2--X***(H)***Y or
Y2--Y***(H)***X
must be between 120-180 degrees. Anything more acute than that is likely
not a bone fide H-bond (but may be a special type of electrostatic
interaction involving pi clouds, etc)
Yes/No?
-Tom
On Tue, Nov 4, 2008 at 2:59 PM, Robert Hanson [EMAIL PROTECTED] wrote:
On Tue, Nov 4, 2008 at 2:23 PM, Thomas Stout [EMAIL PROTECTED]wrote:
The basic premise was: if two appropriate heteroatoms are spatially
located within a set range of separations (1.8-3.4A)
whatever the numbers are, that's trivial, right?
connect 1.8 3.4 {xxx} {yyy}
(Using Jmol script as a stand-in for real code -- not suggesting this will
be a Jmol script in the end.)
and disposed to one another within a certain angular spread (120-180
degrees) relative to the atoms to which they are covalently bonded
There's the more difficult analysis. Can you define that better? Relative
to which atoms? What exactly is the angle being measured? Do you include
only heavy atoms there, or do we need to add H atoms? Is it an average, a
minimum, or a maximum?
I understand the basic idea - that if we had the H atom, we could check to
see if the angle X -- H -- Y was within a certain range. Jmol does that for
assigning the main-chain hydrogen bonds already. (Jmol calculates a position
for the NH hydrogen, does a little calculation to estimate the bond
energy, and from that decides whether or not to assign a hydrogen bond
between groups.)
I'm assuming here we are not only talking about peptide bond NH hydrogens
here (they are trivial to place) -- it's the OH and other NH hydrogens that
could be problematic.
So without the H atom we do something else -- involving the other atoms to
which X and Y are covalently bonded Keep talking.
Just to summarize:
1) If the hydrogens are there, it's easy. For models containing a full set
of H atoms, we can just about do this with a Jmol script already. Make a
hydrogen bond whenever there is an H connected to (O, N, S) within xx
angstroms of (O, N, S) and having X--H--Y angle between a and b -- that's
pretty much just a measure command tied to a connect command.
2) When H atoms are not there -- typical PDB files -- that's where the
difficulty begins.
Certainly an interesting challenge.
You define it well, and I'll implement it.
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Possibly. Is that documented or discussed somewhere in the literature?
I could probably implement that tonight while watching the election returns
then let you try it out. I propose:
calculate hbonds {atomset1} {atomset2}
maybe with some settings for minimum and maximum distances.
Bob
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
I'd have to go dig in the literature for a referenceit's just my
personal guideline(s). I'll see what I can find -- most of that literature
is pretty old
-Tom
On Tue, Nov 4, 2008 at 3:23 PM, Robert Hanson [EMAIL PROTECTED] wrote:
Possibly. Is that documented or discussed somewhere in the literature?
I could probably implement that tonight while watching the election returns
then let you try it out. I propose:
calculate hbonds {atomset1} {atomset2}
maybe with some settings for minimum and maximum distances.
Bob
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
-
This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
Build the coolest Linux based applications with Moblin SDK win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/___
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
Re: [Jmol-users] Protein Hydrogen Bonds again
Thanks for the response to my message. I've been tied up with something and now it's quite late in Europe so I'll just send a brief reply to a couple of points and any further info later. 1. My interest is only in proteins where the hydrogens have been added. The database of the proteins I wish to display is predominantly the Richardson 500, where the resolution is never more than 2 angstroms, all the hydrogens have been added, and ambiguous residues checked manually. We are only interested in displaying H- bonds we are pretty sure about. 2. Our database is of hydrogen-bonded small 3D protein motifs, and includes all the H-bonds in the proteins. These have been derived using Janet Thornton's HBPlus program. However this is not public domain. The program program from the student many years ago is first rate. It was initially written by a lecturer in computing science here in Glasgow and the student made modifications. This was the brightest PhD student my colleague ever had. I don't think there is any question about the quality of the program. The suggestion of writing extra scripts seems unattractive at first sight. At the moment, after a search I am allowing a user to fire up Jmol showing the motifs in the context of the protein all generated on the fly. It's tricky enough to generate javascript controls for a particular situation in the context of the pdb file and the residues in the motifs. I haven't done anything with Jmol for two years (my web app uses 10.2) so perhaps I am overestimating the difficulty of the problem. David PS The web application I am using can be found at http:// motif.gla.ac.uk/motif/index.html. To see what I mean: -Motifs in a Protein -eg 1BX4 and Asx Turn -3D -Single Motif Display -Motif 1 -Hydrogen Bonds on (nothing happens) I would like to display the H-bond from Asn to main-chain NH. If you choose Beta Turn for the same protein, the main-chain - main- chain H-bonds can be displayed. ___ Dr. David P. Leader, Faculty of Biomedical Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK Phone: +44 (0)141 330 5905 http://doolittle.ibls.gla.ac.uk/leader The University of Glasgow, charity number SC004401 ___ - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100url=/ ___ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users
