Dear all,
Thanks a lot for your replies, I will now explain the protocol I followed in order you to completely understand my problem:
( For a 300 DPPC lipid membrane with no proteins):
1. Converting the pre-equilibrated, solvated and neutralised DPPC membrane using the script charmmlipid2amber and the reference table charmmlipid2amber.csv (after this steps the each lipid is a single residue with a TER card after each one).
2. Obtaining box dimensions using the script vmd_box_dims.sh.
3. Loading the system into tleap in order to obtain amber parameters. Setting box dimensions using the values obtained in the step above. ( After saving pdb in this step it is observed that each lipid got divided into 3 residues with a TER card after each one, however it seems to have bonds between residues.)
4. EM: 20000 steps (7500 steepest descent), ntb=1. (After this step pdb structure of the system was obtained using ambpdb and it is observed that each lipid got divided into 3 residues with a TER card after each one)
5. Heating 1: Up to 100 K in 2500 steps holding lipid fixed with soft restraints (10.0). NTV. (ntb=1)
6. Heating 2: Up to 325 K (temperature recommended in references for a DPPC system) in 50000 steps holding lipid fixed with soft restraints (10.0). NTP (ntb= 2)
(After heating pdb was obtained using cpptraj it is observed that each lipid got divided into 3 residues with a TER card after each one. The structure of the lipid seem to be a whole, heads and tails remain together).
7. Hold: 5ns MD NTP in 10 steps and skinnb=5. The system resists the whole simulation, the simulation doesnīt stop running. However lipids get very close to each other forming a very compact and small structure in the centre of the box in which tails and heads are separated and the solvent expands a lot leaving huge gaps in between molecules. (My supervisor described it as the system was blowing up).
NB: Here itīs been described the protocol followed for only the membrane but i have got the same problem when i do it with the anchored protein and when i heat the system up to 303 K in less steps).
Many many thanks for your help.
Sincerely
Raquel
________________________________________
From: Benjamin D Madej <bmadej.ucsd.edu>
Sent: 27 January 2015 00:50
To: AMBER Mailing List
Subject: Re: [AMBER] Lipid 14 format problem
Hi Romero,
First thing to check is the structure in Xleap: Is there a bond drawn between the tail and head groups in each lipid?
Second thing is to check the PDB you load in Leap to build your Amber inputs: Does it have TER lines after each residue or after each group of three residues? There should only be a TER line after each lipid (i.e. each three lipid residues). If all is good with the residue name, residue order, TER lines, and atom names, Leap should automatically build the lipid molecule. There could be a PDB formatting issue as well (i.e. data in the wrong columns).
Third thing to check is periodic box dimensions (volume) vs. time for your initial simulation: Does it decrease rapidly? Is the energy of the system normal? If so, I would probably stick with skinnb=5, and then just restart several times until the box dimensions equilibrate. It is likely that the periodic box is shrinking a lot in a very short time.
Best,
Ben Madej
________________________________________
From: Romero, Raquel [raquel.romero.12.ucl.ac.uk]
Sent: Monday, January 26, 2015 2:07 PM
To: AMBER Mailing List
Subject: Re: [AMBER] Lipid 14 format problem
Dear Benjamin,
Thanks for your reply.
How do you want me to reach the equilibration stage without saving amber parameters? The problem is in both pdb files when i save it using savepdb in tleap and when i obtain in from the prmtop using ambpdb or cpptraj.
I used a skinnb=5 which is the value recommended in the tutorial, and the system blows up. Shall I increase it more? (AS i Said before i followed the tutorial quite faithfully, except for the modifications required for a DPPC system).
I think that the problem is to be with the format as heads and tails go apart during the simulation, detail i forgot to mention before.
Thanks in advance to anyone.
Raquel
________________________________________
From: Benjamin D Madej <bmadej.ucsd.edu>
Sent: 26 January 2015 19:45
To: AMBER Mailing List
Subject: Re: [AMBER] Lipid 14 format problem
1. "tleap divides each lipid residue into three" -- when you look at a DPPC molecule with the xleap "edit" command, are the three residues actually bonded together? If so, then the residue connections are defined.
2. "place a TER card after each one of the new residues" -- is that from a savepdb command in Leap? Or is it after running charmmlipid2amber.py? I suspect it is from the savepdb command.
An important note about Leap,
- Leap reads Lipid14 format PDB files fine. However, Leap *does not write* PDB files in the Lipid14 format. Instead, Leap adds TER lines after every lipid residue.
- There is a workaround to *write* a Lipid14 PDB file from Leap. First, save an Amber prmtop and inpcrd with "saveamberparm", and then use cpptraj to convert to a PDB with "trajin" and "trajout".
I suspect the real problem here is the infamous GPU box dimensions issue. I recommend either increasing the "skinnb" value, or restarting short simulations, or both. I will not go into the technical details here of why this is the case, as I believe it has been discussed elsewhere before. We use this approach to equilibrate box dimensions of bilayers when the density has not converged yet. This is especially an issue with CHARMM-GUI input structures.
Best,
Ben Madej
________________________________________
From: Romero, Raquel [raquel.romero.12.ucl.ac.uk]
Sent: Monday, January 26, 2015 8:56 AM
To: AMBER Mailing List
Subject: Re: [AMBER] Lipid 14 format problem
Hi Ross,
Thanks a lot for your reply. This is the output i got:
> pwd
/usr/local/amber14
> ./update_amber --show-applied-patches
AmberTools 14 Applied Patches:
------------------------------
update.1, update.2, update.3, update.4, update.5, update.6, update.7, update.8, update.9, update.10,
update.11, update.12, update.13, update.14, update.15, update.16, update.17, update.18, update.19, update.20,
update.21, update.22
Amber 14 Applied Patches:
-------------------------
update.1 (modifies pmemd.amoeba)
update.2 (modifies pmemd)
update.3 (modifies pmemd, pmemd.cuda)
update.4 (modifies pmemd)
update.5 (modifies pmemd)
update.6 (modifies pmemd.cuda, pmemd.cuda.MPI)
update.7 (modifies pmemd, pmemd.MPI)
update.8 (modifies pmemd, pmemd.MPI)
> ls -la $AMBERHOME/bin/*leap*
-rwxr-xr-x 1 root root 10185 Dec 2 12:22 /usr/local/amber14/bin/pytleap
-rwxr-xr-x 1 root root 341 Dec 2 12:20 /usr/local/amber14/bin/tleap
-rwxr-xr-x 1 root root 406 Dec 2 12:20 /usr/local/amber14/bin/xleap
Best wishes
Many thanks
Raquel
________________________________________
From: Ross Walker <ross.rosswalker.co.uk>
Sent: 26 January 2015 15:18
To: AMBER Mailing List
Subject: Re: [AMBER] Lipid 14 format problem
Hi Raquel,
This sounds like a bug in leap that may have been introduced by a recent update - indeed someone in my group reported the same behavior to me earlier this week with our lab copy of AMBER 14 but then switched to our development version and it worked fine so this deserves further investigation. First can you let us know what version of AmberTools and leap you are using. Do the following and copy the output here please:
cd $AMBERHOME
pwd
./update_amber --show-applied-patches
ls -la $AMBERHOME/bin/*leap*
Thanks.
All the best
Ross
> On Jan 26, 2015, at 4:44 AM, Romero, Raquel <raquel.romero.12.ucl.ac.uk> wrote:
>
> Dear all,
>
>
> I want to run MD using amber over a system containing a DPPC lipid bilayer and two proteins. After having my bilayer in Charmm format and converting it into a lipid 14 readable format, I have loaded it in tleap. I have observed that, unlike it is described in the tutorials and references, tleap divides each lipid residue into three and place a TER card after each one of the new residues. (The bibliography says each lipid should remain divided into three residues but all three part of a chain). I think that this format is spoiling my simulations. I have followed the protocol described in papers and tutorials (with the pertinent changes for a DPPC system). My system blows up in the first phase of the 5 ns MD with barostat for equilibrating box dimensions.
>
>
> Can anyone please tell me what is the way to obtain the right format of Lipid 14 phospholipids? Or how to get forward with my simulation with this apparently wrong format?
>
>
> Thanks in advance to everyone!
>
>
>
> Raquel Romero
> PhD Candidate
> Department of Pharmaceutical and Biological Chemistry
> UCL School of Pharmacy, University College London
> 29/39 Brunswick Square
> London
> WC1N 1AX
>
> T: 0207 753 5972
> E: raquel.romero.12.ucl.ac.uk
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Jan 27 2015 - 06:00:05 PST
