[gmx-users] ligand falling out of active site during EM

[Diane Fournier]

Hello !

I'm trying to run a molecular dynamics on a drug-enzyme complex. I did John Kerrigan's tutorial and everything worked fine. Now I'm trying with my system but I get a problem : the ligand keeps falling out of the active site during EM. I thought maybe it was a pbc problem and used comm-grps = protein in my .mdp file, but I get the same result. I transformed the .gro input file to .pdb to view it in pymol and the ligand is in the active site before simulation. So it seems this happens during steepest descents EM.

The ligand is a hybrid inhibitor containing a steroid moiety (estradiol) linked to an adenosine-like moiety with a 13-methylene alkyl chain.

Is there a way to keep/force the ligand in the active site during EM (maybe using PR) ?

Is this reflecting some physical phenomenon, ie the ligand has low affinity ?


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060505/319293db/attachment.html>

[David Mobley]

Hi,

Post by Diane Fournier
I'm trying to run a molecular dynamics on a drug-enzyme complex. I did John
Kerrigan's tutorial and everything worked fine. Now I'm trying with my
system but I get a problem : the ligand keeps falling out of the active site
during EM. I thought maybe it was a pbc problem and used comm-grps = protein
in my .mdp file, but I get the same result. I transformed the .gro input
file to .pdb to view it in pymol and the ligand is in the active site before
simulation. So it seems this happens during steepest descents EM.

This means that it is lower energy for the ligand not to be in the
binding site (at least, in its starting configuration) for some
reason, assuming you have everything set up properly.

Post by Diane Fournier
Is there a way to keep/force the ligand in the active site during EM (maybe using PR) ?

You could restrain it to stay in the binding site using distance
restraints or position restraints or some such. But this may only mask
some larger problem.

Post by Diane Fournier
Is this reflecting some physical phenomenon, ie the ligand has low affinity ?

Hard to say. Where are you getting your starting coordinates? I doubt
it is reflecting low affinity, as in my experience even low affinity
ligands will stably stay in binding sites for relatively long times in
MD, and definitely during energy minimization. It more likely means
that your starting structure is not very good -- i.e. the ligand is
sterically clashing with something in the binding site such that
energy minimization moves the ligand to lower energy by removing the
steric clashes and partly pushing the ligand out of the binding site.

There are of course other possibilities... Your ligand parameters
could be bad, etc.

You could certainly solve the problem of it leaving during energy
minimization by restraining it to stay in the binding site, but then
it will probably just leave during MD, so I think you'd probably just
be masking the problem. I would look carefully at your starting
structure and doublecheck your parameters and so on.

David

Post by Diane Fournier
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read
http://www.gromacs.org/mailing_lists/users.php

[Diane Fournier]

Still on the same problem, I made a pr run on the complex, and had the same result (ligand is out of the active site at time = 0.000 ps. Then I ran the same pr run, but with dt = 0.001 ps with all coordinates output for my trajectory. It turns out the ligand starts out of the active site, even if my input coordinates have the ligand inside. What is happening ??


-----Original Message-----
From: gmx-users-bounces at gromacs.org on behalf of Diane Fournier
Sent: Fri 5/5/2006 3:05 PM
To: gmx-users at gromacs.org
Subject: [gmx-users] ligand falling out of active site during EM

Hello !

I'm trying to run a molecular dynamics on a drug-enzyme complex. I did John Kerrigan's tutorial and everything worked fine. Now I'm trying with my system but I get a problem : the ligand keeps falling out of the active site during EM. I thought maybe it was a pbc problem and used comm-grps = protein in my .mdp file, but I get the same result. I transformed the .gro input file to .pdb to view it in pymol and the ligand is in the active site before simulation. So it seems this happens during steepest descents EM.

The ligand is a hybrid inhibitor containing a steroid moiety (estradiol) linked to an adenosine-like moiety with a 13-methylene alkyl chain.

Is there a way to keep/force the ligand in the active site during EM (maybe using PR) ?

Is this reflecting some physical phenomenon, ie the ligand has low affinity ?



-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 3432 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060508/1b2c59db/attachment.bin>

[David Mobley]

Diane,

Post by Diane Fournier
Still on the same problem, I made a pr run on the complex, and had the same result (ligand is out of the active site at time = 0.000 ps. Then I ran the same pr run, but with dt = 0.001 ps with all coordinates output for my trajectory. It turns out the ligand starts out of the active site, even if my input coordinates have the ligand inside. What is happening ??

First, if you are doing energy minimization, dt does nothing...

But second, double-check where it starts. For example, generate a tpr
file and then use trjconv to convert your initial gro file to a pdb
file and open it up with some viewer (i.e. pymol) (Or I guess use vmd
to visualize your starting gro file). See if the ligand starts outside
the binding site.

If so, then it probably means you've done something in a funny order.
For example, if you use editconf and genbox on just the protein before
adding the ligand, well, those tools re-center teh protein in the box,
so it will get shifted relative to the ligand (and hence the ligand
will no longer be in the binding site). I think the solution to that
is to use editconf or genbox on the system (protein+ligand), not just
the protein. If that isn't the problem, it might be worth e-mailing
the list exactly the steps you're following, and double-checking that
the initial protein and ligand coordinates prior to setting up the
system in GROMACS have the ligand in the binding site.

David

Post by Diane Fournier
-----Original Message-----
From: gmx-users-bounces at gromacs.org on behalf of Diane Fournier
Sent: Fri 5/5/2006 3:05 PM
To: gmx-users at gromacs.org
Subject: [gmx-users] ligand falling out of active site during EM
Hello !
I'm trying to run a molecular dynamics on a drug-enzyme complex. I did John Kerrigan's tutorial and everything worked fine. Now I'm trying with my system but I get a problem : the ligand keeps falling out of the active site during EM. I thought maybe it was a pbc problem and used comm-grps = protein in my .mdp file, but I get the same result. I transformed the .gro input file to .pdb to view it in pymol and the ligand is in the active site before simulation. So it seems this happens during steepest descents EM.
The ligand is a hybrid inhibitor containing a steroid moiety (estradiol) linked to an adenosine-like moiety with a 13-methylene alkyl chain.
Is there a way to keep/force the ligand in the active site during EM (maybe using PR) ?
Is this reflecting some physical phenomenon, ie the ligand has low affinity ?
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

[Arneh Babakhani]

Hello all,

How do you remove unwanted waters (or how do you prevent genbox from
placing unwanted waters) after you solvate by genbox?

This problem was posted once before,
http://www.gromacs.org/pipermail/gmx-users/2005-December/018997.html

but it doesn't seem like it was resolved. Is there an easy way to do
this (aside from going into the .gro file and manually deleting the
unwanted waters!)

Thanks,

Arneh

[David van der Spoel]

Post by Arneh Babakhani
Hello all,
How do you remove unwanted waters (or how do you prevent genbox from
placing unwanted waters) after you solvate by genbox?
This problem was posted once before,
http://www.gromacs.org/pipermail/gmx-users/2005-December/018997.html
but it doesn't seem like it was resolved. Is there an easy way to do
this (aside from going into the .gro file and manually deleting the
unwanted waters!)

genbox 3.3.1 has an option to set the maximum number of waters.

Post by Arneh Babakhani
Thanks,
Arneh
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

[Arneh Babakhani]

Thanks for the input. Is there a way to restrict the water from a
certain space (in other words, to confine it a certain location)?

Thanks,

Arneh

Post by David van der Spoel

Post by Arneh Babakhani
Hello all,
How do you remove unwanted waters (or how do you prevent genbox from
placing unwanted waters) after you solvate by genbox?
This problem was posted once before,
http://www.gromacs.org/pipermail/gmx-users/2005-December/018997.html
but it doesn't seem like it was resolved. Is there an easy way to do
this (aside from going into the .gro file and manually deleting the
unwanted waters!)

genbox 3.3.1 has an option to set the maximum number of waters.

Post by Arneh Babakhani
Thanks,
Arneh
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

[David van der Spoel]

Post by Arneh Babakhani
Thanks for the input. Is there a way to restrict the water from a
certain space (in other words, to confine it a certain location)?

just the shell option, otherwise no.

genbox -h

Post by Arneh Babakhani
Thanks,
Arneh

Post by David van der Spoel

Post by Arneh Babakhani
Hello all,
How do you remove unwanted waters (or how do you prevent genbox from
placing unwanted waters) after you solvate by genbox?
This problem was posted once before,
http://www.gromacs.org/pipermail/gmx-users/2005-December/018997.html
but it doesn't seem like it was resolved. Is there an easy way to do
this (aside from going into the .gro file and manually deleting the
unwanted waters!)

genbox 3.3.1 has an option to set the maximum number of waters.

Post by Arneh Babakhani
Thanks,
Arneh
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

[Diane Fournier]

Thank you !

I think the problem was indeed with building the box, because I redid the whole sequence on my drug-enzyme system (building the box with editconf, putting the water with genbox, and then writing my .tpr file with grompp) and ran the same (test) position restraint md run, and this time, the ligand was inside. Will try minimisation again.

Diane


-----Original Message-----
From: gmx-users-bounces at gromacs.org on behalf of David Mobley
Sent: Mon 5/8/2006 12:22 PM
To: Discussion list for GROMACS users
Subject: Re: [gmx-users] ligand falling out of active site during EM

Diane,

Post by Diane Fournier
Still on the same problem, I made a pr run on the complex, and had the same result (ligand is out of the active site at time = 0.000 ps. Then I ran the same pr run, but with dt = 0.001 ps with all coordinates output for my trajectory. It turns out the ligand starts out of the active site, even if my input coordinates have the ligand inside. What is happening ??

First, if you are doing energy minimization, dt does nothing...

But second, double-check where it starts. For example, generate a tpr
file and then use trjconv to convert your initial gro file to a pdb
file and open it up with some viewer (i.e. pymol) (Or I guess use vmd
to visualize your starting gro file). See if the ligand starts outside
the binding site.

If so, then it probably means you've done something in a funny order.
For example, if you use editconf and genbox on just the protein before
adding the ligand, well, those tools re-center teh protein in the box,
so it will get shifted relative to the ligand (and hence the ligand
will no longer be in the binding site). I think the solution to that
is to use editconf or genbox on the system (protein+ligand), not just
the protein. If that isn't the problem, it might be worth e-mailing
the list exactly the steps you're following, and double-checking that
the initial protein and ligand coordinates prior to setting up the
system in GROMACS have the ligand in the binding site.

David

Post by Diane Fournier
-----Original Message-----
From: gmx-users-bounces at gromacs.org on behalf of Diane Fournier
Sent: Fri 5/5/2006 3:05 PM
To: gmx-users at gromacs.org
Subject: [gmx-users] ligand falling out of active site during EM
Hello !
I'm trying to run a molecular dynamics on a drug-enzyme complex. I did John Kerrigan's tutorial and everything worked fine. Now I'm trying with my system but I get a problem : the ligand keeps falling out of the active site during EM. I thought maybe it was a pbc problem and used comm-grps = protein in my .mdp file, but I get the same result. I transformed the .gro input file to .pdb to view it in pymol and the ligand is in the active site before simulation. So it seems this happens during steepest descents EM.
The ligand is a hybrid inhibitor containing a steroid moiety (estradiol) linked to an adenosine-like moiety with a 13-methylene alkyl chain.
Is there a way to keep/force the ligand in the active site during EM (maybe using PR) ?
Is this reflecting some physical phenomenon, ie the ligand has low affinity ?
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 4684 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060508/0a1f731a/attachment.bin>

[Anton Feenstra]

Post by Diane Fournier
Thank you !
I think the problem was indeed with building the box, because I redid
the whole sequence on my drug-enzyme system (building the box with
editconf, putting the water with genbox, and then writing my .tpr
file with grompp) and ran the same (test) position restraint md run,
and this time, the ligand was inside. Will try minimisation again.

For the record: you have almost certainly had problems with PBC. mdrun
will always place molecules 'in the box' (see manual). If your complex
happens to be on the box edge, one molecule may end up at a different
side of the box.

If, instead, your complex was in the middle of the box, as probably
happend during your second setup sequence, pbc will not affect it, at
least not during EM when molecules tend not to move much.

Note, that for different box shapes, the 'middle of the box' may not be
what you expect.
--
Groetjes,

Anton

* NOTE: New Affiliation, Phone & Fax numbers (below) *
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | IBIVU/Bioinformatics - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081A HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel [phone] - Fax [phone] - Room P440 |
| | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" (RHCP) |
|_____________|_______________________________________________________|

[Tsjerk Wassenaar]

Hi,

Maybe as a note for any interested but unaware. In gromacs, the middle of
the box is always the middle of the *rectangular* box defined by the first
three numbers in the last line of the .gro file.

Cheers,

Tsjerk

Post by Anton Feenstra

Post by Diane Fournier
Thank you !
I think the problem was indeed with building the box, because I redid
the whole sequence on my drug-enzyme system (building the box with
editconf, putting the water with genbox, and then writing my .tpr
file with grompp) and ran the same (test) position restraint md run,
and this time, the ligand was inside. Will try minimisation again.

For the record: you have almost certainly had problems with PBC. mdrun
will always place molecules 'in the box' (see manual). If your complex
happens to be on the box edge, one molecule may end up at a different
side of the box.
If, instead, your complex was in the middle of the box, as probably
happend during your second setup sequence, pbc will not affect it, at
least not during EM when molecules tend not to move much.
Note, that for different box shapes, the 'middle of the box' may not be
what you expect.
--
Groetjes,
Anton
* NOTE: New Affiliation, Phone & Fax numbers (below) *
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | IBIVU/Bioinformatics - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081A HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel [phone] - Fax [phone] - Room P440 |
| | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" (RHCP) |
|_____________|_______________________________________________________|
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

--

Tsjerk A. Wassenaar, M.Sc.
Groningen Biomolecular Sciences and Biotechnology Institute (GBB)
Dept. of Biophysical Chemistry
University of Groningen
Nijenborgh 4
9747AG Groningen, The Netherlands
[phone]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060510/e07d49e2/attachment.html>

[chris.neale]

If I have a solvated membrane and I want to extend the z dimension and add new
waters there, this is what I do. The script could be modified to get rid of new
waters in any x/y/z dimensions (around line 41 of the script).

1. run genbox on initial.gro to create solvated.gro
2. cp solvated.gro new_waters.gro
3. use vi to remove everything in new_waters.gro except the new waters (make
sure you remove waters that were in initial.gro)
4. use vi to edit keepbyz.pl
- upperz and lowerz variables as you please
- sol to the name of your solvent molecule
5. run keepbyz.pl on new_waters.gro
./keepbyz new_waters.gro > keep_these_waters.gro
6. tail -1 initial.gro > last_line.gro
7. head -$(expr $(cat initial.gro | wc -l | awk '{print $1}') - 1 ) initial.gro

not_last_line.gro

8. cat not_last_line.gro new_waters.gro last_line.gro > new_system.gro
9. editconf -f new_system.gro -o new_system_sequential_numbers.gro

Here is the script (make sure you chmod +x keepbyz.pl):

#!/bin/bash
# give x.gro as first command line arguement
upperz=6.417
lowerz=0.820
sol=SOL
count=0
cat $1 | grep "$sol" | while read line; do
for first in $line; do
break
done
if [ "$count" = 3 ]; then
count=0
fi
count=$(expr $count + 1)
if [ "$count" != 1 ]; then
continue
fi
l=${#line}
m=$(expr $l - 24)
i=1
for word in ${line:$m}; do
if [ "$i" = 1 ]; then
popex=$word
else
if [ "$i" = 2 ]; then
popey=$word
else
if [ "$i" = 3 ]; then
popez=$word
break
fi
fi
fi
i=$(expr $i + 1)
done
nolx=`echo "$popez > $upperz" | bc`
nohx=`echo "$popez < $lowerz" | bc`
myno=$(expr $nolx + $nohx)
if [ "$myno" != 0 ]; then
z=${#first}
if [ "$z" != 8 ]; then
sfirst="[[:space:]]$first"
else
sfirst=$first
fi
`echo grep $sfirst $1`
fi
done

[Arneh Babakhani]

Great, thank you very much, will try it out,

Post by chris.neale
If I have a solvated membrane and I want to extend the z dimension and add new
waters there, this is what I do. The script could be modified to get rid of new
waters in any x/y/z dimensions (around line 41 of the script).
1. run genbox on initial.gro to create solvated.gro
2. cp solvated.gro new_waters.gro
3. use vi to remove everything in new_waters.gro except the new waters (make
sure you remove waters that were in initial.gro)
4. use vi to edit keepbyz.pl
- upperz and lowerz variables as you please
- sol to the name of your solvent molecule
5. run keepbyz.pl on new_waters.gro
./keepbyz new_waters.gro > keep_these_waters.gro
6. tail -1 initial.gro > last_line.gro
7. head -$(expr $(cat initial.gro | wc -l | awk '{print $1}') - 1 ) initial.gro

not_last_line.gro

8. cat not_last_line.gro new_waters.gro last_line.gro > new_system.gro
9. editconf -f new_system.gro -o new_system_sequential_numbers.gro
#!/bin/bash
# give x.gro as first command line arguement
upperz=6.417
lowerz=0.820
sol=SOL
count=0
cat $1 | grep "$sol" | while read line; do
for first in $line; do
break
done
if [ "$count" = 3 ]; then
count=0
fi
count=$(expr $count + 1)
if [ "$count" != 1 ]; then
continue
fi
l=${#line}
m=$(expr $l - 24)
i=1
for word in ${line:$m}; do
if [ "$i" = 1 ]; then
popex=$word
else
if [ "$i" = 2 ]; then
popey=$word
else
if [ "$i" = 3 ]; then
popez=$word
break
fi
fi
fi
i=$(expr $i + 1)
done
nolx=`echo "$popez > $upperz" | bc`
nohx=`echo "$popez < $lowerz" | bc`
myno=$(expr $nolx + $nohx)
if [ "$myno" != 0 ]; then
z=${#first}
if [ "$z" != 8 ]; then
sfirst="[[:space:]]$first"
else
sfirst=$first
fi
`echo grep $sfirst $1`
fi
done
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php