Callum, can you show Subbarao how you are getting DMPC at 56A^2 - I believe
the 50A^2 being seen here is an issue with protocol - essentially if you are
too aggressive you can freeze your bilayer and then it will take a LONG
(read 10s of microseconds) to 'unfreeze'. So I think the issue is more one
of simulation protocol here than the GAFF Lipid parameters.
From: Subbarao Kanchi <ksubbuamber.gmail.com>
Date: Wednesday, October 9, 2013 1:44 AM
To: "Dickson, Callum" <callum.dickson09.imperial.ac.uk>
Cc: AMBER Mailing List <amber.ambermd.org>, Ross Walker
<ross.rosswalker.co.uk>
Subject: Re: [AMBER] Fwd: DMPC GAFF lipid membrane is freezing at 303 K.
Dear Callum,
Thank you for your kind reply. I tested the GAFFLipid
force field extensively on CPUs. The area per lipid is going even less than
56 A^2 and bilayer getting freeze.I got 49 A^2 after 50 ns. I hope this will
be fixed in next coming amber force filed (Lipid 12).
thanking you all
Regards,
Subbarao Kanchi
On Wed, Oct 9, 2013 at 1:37 PM, Dickson, Callum
<callum.dickson09.imperial.ac.uk> wrote:
> Hi Subbarao,
>
> We have had a number of issues with bilayer simulations in amber related to
> the rapid collapse of the simulation box in the initial stages of simulation.
> These were a particular issue when using the GPU code. This affected the
> gafflipid work and mean the lipid parameters require more attention.
>
> As a consequence running a DMPC with gafflipid causes the area per lipid to
> drop to about 56A^2, which is below the experimental 60A^2 though the bilayer
> is still liquid phase.
>
> If you want to run DMPC at the experimental area per lipid using amber the
> following options are open to you:
>
> - use GAFF parameters by Lula Rosso with an applied surface tension (available
> on the Bryce website)
> - use GAFFlipid with a (lower) applied surface tension
> - wait for Lipid12 which will support DMPC and will be released soon
>
> Kind regards,
> Callum
>
>
> From: Subbarao Kanchi [ksubbuamber.gmail.com]
> Sent: 07 October 2013 06:05
> To: AMBER Mailing List
> Cc: Dickson, Callum; Ross Walker
> Subject: Re: [AMBER] Fwd: DMPC GAFF lipid membrane is freezing at 303 K.
>
> Dear Ross,
> I would like to know problem in using the DMPC GAFFLipid force
> field.I already attached my Amber input file and I would appreciate any
> suggestion from you.
>
>
> Regards,
> Subbarao Kanchi
>
>
> On Wed, Oct 2, 2013 at 7:55 AM, Subbarao Kanchi <ksubbuamber.gmail.com> wrote:
>>
>> Dear Ross,
>> I follow the following protocol.
>> 1) minimization 10000 (steepest descant) + 5000 (conjugate gradient)
>> 2) Heating (from 0 to 303 K)
>> 3) production run (amber input file is attached md.in <http://md.in> )
>>
>> Making DMPC bilayer system.
>> I have used the same bilayer patch that provided in the Lipid book
>> (288_dmpc.pdb), library (DMP.lib) and parameter file (GAFFLipid.frcmod).
>>
>> the input files are prepared in Xleap
>>
>>
>> source leaprc.ff10
>> source leaprc.gaff
>> source GAFFlipid.dat
>> loadamberparams GAFFlipid.frcmod
>> loadoff DMP.lib
>> lipid = loadpdb 288dmpc_100ns.pdb
>> setBox lipid centers
>> saveamberparm lipid 288dmpc.prmtop 288dmpc.inpcrd
>>
>>
>> the .prmtop and .inpcrd files were used for the amber simulations.
>> The amber version is
>> -------------------------------------------------------
>> Amber 10 SANDER 2008
>> -------------------------------------------------------
>>
>> | PMEMD implementation of SANDER, Release 10
>>
>>
>>
>>
>>
>>
>>
>> On Tue, Oct 1, 2013 at 9:34 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
>>> Hi Subbarao,
>>>
>>> I believe Callum is on vacation right now so it might be a while before
>>> someone has the time to try things out here and find out why you are
>>> seeing problems. In the meantime it would be helpful if you could post
>>> your entire simulation procedure from structure generation all the way to
>>> you seeing it freezing. Also please let us know the specific version of
>>> AMBER you are using, command lines and the patch level of the code. Lipid
>>> simulations are extremely sensitive to the way the MD is carried out so it
>>> is critical to know this information in order to figure out why you are
>>> seeing problems.
>>>
>>> All the best
>>> Ross
>>>
>>>
>>> On 9/30/13 11:14 AM, "Subbarao Kanchi" <ksubbuamber.gmail.com> wrote:
>>>
>>>> >*Dear Amber users,*
>>>> >* I have used the GAFF Lipid parameters
>>>> >(**provided
>>>> >at the lipid book http://lipidbook.bioch.ox.ac.uk/package/)** for DMPC
>>>> >and
>>>> >not able to get the reported area per lipid (was run number of times). May
>>>> >I ask Authors of GAFFLipid paper (GAFFlipid: a General Amber Force Field
>>>> >for the accurate molecular dynamics simulation of phospholipid) for the
>>>> >correct
>>>> >GAFFLipid parameters. *
>>>> >*
>>>> >*
>>>> >*
>>>> >*
>>>> >*
>>>> >*
>>>> >*Regards,*
>>>> >*Subbarao Kanchi. *
>>>> >
>>>> >
>>>> >---------- Forwarded message ----------
>>>> >From: Subbarao Kanchi <ksubbuamber.gmail.com>
>>>> >Date: Fri, Sep 27, 2013 at 11:52 PM
>>>> >Subject: Re: [AMBER] DMPC GAFF lipid membrane is freezing at 303 K.
>>>> >To: jojartb.jgypk.u-szeged.hu
>>>> >Cc: AMBER Mailing List <amber.ambermd.org>, Ross Walker <
>>>> >ross.rosswalker.co.uk>, callum.dickson09.imperial.ac.uk
>>>> >
>>>> >
>>>> >*Dear callum and Balazs*,
>>>> > *I have performed the DMPC simulation
>>>> >(for 288 lipid patch) with this GAFFLipid parameter (provided at the lipid
>>>> >book http://lipidbook.bioch.ox.ac.uk/package/). I got the area per Lipid
>>>> >~55.22 sq Angstroms/lipid after 10ns but it is far away from the reported
>>>> >value in the paper "GAFFlipid: a General Amber Force Field for the
>>>> >accurate
>>>> >molecular dynamics simulation of phospholipid" (60.38 sq Angsroms/lipid).
>>>> >I
>>>> >would appreciate any suggestion in this regards. *
>>>> >*
>>>> >*
>>>> >*Regards,*
>>>> >*Subbarao kanchi. *
>>>> >
>>>> >
>>>> >On Wed, Sep 25, 2013 at 1:00 AM, Subbarao Kanchi
>>>> ><ksubbuamber.gmail.com>wrote:
>>>> >
>>>>> >> Hi Balazs,
>>>>> >> Thank you for the reply and I am trying the new
>>>>> parameters
>>>>> >> form this Lipidbook link.I will update the results of this new
>>>>> >>parameters.
>>>>> >>
>>>>> >>
>>>>> >> Regards,
>>>>> >> Subbarao Kanchi
>>>>> >>
>>>>> >>
>>>>> >> On Tue, Sep 24, 2013 at 11:32 PM, <jojartb.jgypk.u-szeged.hu> wrote:
>>>>> >>
>>>>>> >>> Dear Subbaro,
>>>>>> >>> The parameters of that website is not GAFFLipid.
>>>>>> >>> Download the correct parameters and coordinate files from
>>>>>> >>> http://lipidbook.bioch.ox.ac.**uk/
>>>>>> <http://lipidbook.bioch.ox.ac.uk/>.
>>>>>> >>> Hope this helps.
>>>>>> >>> Best regards, Balazs
>>>>>> >>>
>>>>>> >>> Idézet (Subbarao Kanchi <ksubbuamber.gmail.com>):
>>>>>> >>>
>>>>>> >>> Hi Ross,
>>>>>>> >>>>
>>>>>>> >>>> Strong apologies for multiple time post. It was happen because
>>>>>>> there
>>>>>>> >>>>is
>>>>>>> >>>> problem to my amber mailing-list account setting and I was not sure
>>>>>>> >>>>that
>>>>>>> >>>> the mail was posted.
>>>>>>> >>>>
>>>>>>> >>>> I simply changed the mass number and I do not get the new VDW
>>>>>>> >>>>parameters.
>>>>>>> >>>> Callum mentioned that the rest of uploaded parameters are correct.
I
>>>>>>> >>>>am
>>>>>>> >>>> fallowing the same protocol and input parameters (same thermostat
and
>>>>>>> >>>> pressure couplings etc) that are mentioned in the force field paper
(
>>>>>>> >>>> GAFFlipid: a General Amber Force Field for the accurate molecular
>>>>>>> >>>> dynamics
>>>>>>> >>>> simulation of phospholipid ) and the parameters that I downloaded
are
>>>>>>> >>>> from
>>>>>>> >>>> this link
>>>>>>> >>>>http://www.pharmacy.**manchester.ac.uk/bryce/amber
>>>>>>> <http://manchester.ac.uk/bryce/amber> <http://www.pharmacy.
>>>>>>> >>>>manchester.ac.uk/bryce/amber <http://manchester.ac.uk/bryce/amber> >
>>>>>>> >>>> .
>>>>>>> >>>>
>>>>>>> >>>> With regards,
>>>>>>> >>>> Subbarao Kanchi.
>>>>>>> >>>>
>>>>>>> >>>>
>>>>>>> >>>>
>>>>>>> >>>>
>>>>>>> >>>> On Tue, Sep 24, 2013 at 10:05 PM, Ross Walker
>>>>>>> <ross.rosswalker.co.uk>
>>>>>>> >>>> wrote:
>>>>>>> >>>>
>>>>>>> >>>> Hi Subbarao,
>>>>>>>> >>>>>
>>>>>>>> >>>>> I am not entire sure why you posted this question multiple times
to
>>>>>>>> >>>>>the
>>>>>>>> >>>>> mailing list. Often it can take several days before people find
the
>>>>>>>> >>>>> time to
>>>>>>>> >>>>> reply.
>>>>>>>> >>>>>
>>>>>>>> >>>>> I would start by asking whether you simply just changed the H
>>>>>>>> mass or
>>>>>>>> >>>>> you
>>>>>>>> >>>>> actually went and got the new VDW parameters as well. What was
the
>>>>>>>> >>>>>exact
>>>>>>>> >>>>> procedure you used - with out knowing this it is difficult to
know
>>>>>>>> >>>>> specifically parameters you used.
>>>>>>>> >>>>>
>>>>>>>> >>>>> All the best
>>>>>>>> >>>>> Ross
>>>>>>>> >>>>>
>>>>>>>> >>>>> From: Subbarao Kanchi <ksubbuamber.gmail.com>
>>>>>>>> >>>>> Date: Tuesday, September 24, 2013 1:56 AM
>>>>>>>> >>>>> To:
>>>>>>>>
>>>>><callum.dickson09.imperial.ac.**uk<callum.dickson09.imperial.ac.uk>>,
>>>>>>>> >>>>> Ross Walker <ross.rosswalker.co.uk>
>>>>>>>> >>>>> Cc: <amber.ambermd.org>
>>>>>>>> >>>>> Subject: Re: DMPC GAFF lipid membrane is freezing at 303 K.
>>>>>>>> >>>>>
>>>>>>>> >>>>> Hi Callum and Ross,
>>>>>>>> >>>>>
>>>>>>>> >>>>> Thank you for your reply. I did not
use
>>>>>>>> >>>>>any
>>>>>>>>> >>>>>> amber GPU version. I used only the amber 10 CPU version.As you
both
>>>>>>>>> >>>>>> suggested, I corrected the hydrogen mass from 1.088 to 1.088 in
>>>>>>>>> >>>>>> GAFFlipid.frcmod file and rerun the simulation still the DMPC
>>>>>>>>> >>>>>>bilayer
>>>>>>>>> >>>>>> is
>>>>>>>>> >>>>>> freezing at 303K. It will be great if you provide correct
>>>>>>>>> parameters
>>>>>>>>> >>>>>> and I
>>>>>>>>> >>>>>> would appreciate receiving any suggestion in this regard.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> With regards,
>>>>>>>>> >>>>>> Subbarao Kanchi.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> Hi Subbarao,
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> So the hl mass should be 1.008. The parameters on Lipidbook
have
>>>>>>>>> >>>>>>been
>>>>>>>>> >>>>>> updated
>>>>>>>>> >>>>>> and I will see that the Bryce website is also updated.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> As Ross mentioned there previously was a bug in the GPU code
that
>>>>>>>>> >>>>>>gave
>>>>>>>>> >>>>>> us
>>>>>>>>> >>>>>> issues with bilayer simulations, including the gafflipid work.
Just
>>>>>>>>> >>>>>> now we
>>>>>>>>> >>>>>> are working to update the Lipid11 parameters, this will include
>>>>>>>>> >>>>>> support for
>>>>>>>>> >>>>>> DMPC bilayers.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> Callum
>>>>>>>>> >>>>>> ______________________________**__________
>>>>>>>>> >>>>>> From: Ross Walker [ross.rosswalker.co.uk
>>>>>>>>> <http://ross.rosswalker.co.uk> ]
>>>>>>>>> >>>>>> Sent: 21 August 2013 18:42
>>>>>>>>> >>>>>> To: AMBER Mailing List
>>>>>>>>> >>>>>> Cc: Dickson, Callum
>>>>>>>>> >>>>>> Subject: Re: [AMBER] DMPC GAFF lipid membrane is freezing at
303 K.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> Hi Subbarao,
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> A couple of things.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> 1) Can you confirm which version of AMBER you are using to run
these
>>>>>>>>> >>>>>> simulations and if it is fully patched. In particular if you
are
>>>>>>>>> >>>>>>using
>>>>>>>>> >>>>>> the
>>>>>>>>> >>>>>> GPU version make sure it is patched to at least v12.3 (bug
>>>>>>>>> fix.18)
>>>>>>>>> >>>>>> since
>>>>>>>>> >>>>>> there were a number of bugs discovered related to pressure
>>>>>>>>> control
>>>>>>>>> >>>>>>in
>>>>>>>>> >>>>>> lipid simulations - especially in cases where the initial box
size
>>>>>>>>> >>>>>> changes
>>>>>>>>> >>>>>> considerably. This bug appeared if the box shrunk too much and
if it
>>>>>>>>> >>>>>> did
>>>>>>>>> >>>>>> would create a feedback loop leading to the box shrinking
>>>>>>>>> further
>>>>>>>>> >>>>>> giving
>>>>>>>>> >>>>>> you a gel phase.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> 2) I just checked the parameters from the website you list and
they
>>>>>>>>> >>>>>> appear
>>>>>>>>> >>>>>> to have not been properly updated.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> Specifically the frcmod file appears to still have the
>>>>>>>>> incorrect H
>>>>>>>>> >>>>>>mass
>>>>>>>>> >>>>>> (and thus H vow params) included:
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> g3 12.010 0.878 same as c3
>>>>>>>>> >>>>>> hl 1.088 0.135 same as hc
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> I am cc'ing Callum here who can verify for you what the new
>>>>>>>>> >>>>>>parameters
>>>>>>>>> >>>>>> should be and where to download them from. Callum can you also
make
>>>>>>>>> >>>>>> sure
>>>>>>>>> >>>>>> the manchester website is updated please.
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> All the best
>>>>>>>>> >>>>>> Ross
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> On 8/21/13 6:55 AM, "Subbarao Kanchi" <ksubbuamber.gmail.com
>>>>>>>>> <http://ksubbuamber.gmail.com> > wrote:
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>> *>Dear Amber users, *
>>>>>>>>> >>>>>> *> I am trying to reproduce the* DMPC* lipid membrane *
>>>>>>>>> >>>>>> *>simulation *
>>>>>>>>> >>>>>> *>using* GaffLipid* force field at the temperature *303* K as
>>>>>>>>> >>>>>>reported
>>>>>>>>> >>>>>> in *
>>>>>>>>> >>>>>> *>the *
>>>>>>>>> >>>>>> *>paper (*GAFFlipid: a General Amber Force Field for the
>>>>>>>>> accurate
>>>>>>>>> >>>>>> molecular *
>>>>>>>>> >>>>>> *>dynamics simulation of phospholipid* ). I used the
>>>>>>>>> equilibrated
>>>>>>>>> >>>>>>DMPC
>>>>>>>>> >>>>>> *
>>>>>>>>> >>>>>> *>membrane (*288dmpc_100ns.pdb,* *DMP.lib,GAFFlipid.frcmod,
>>>>>>>>> >>>>>> GAFFlipid.dat*) *
>>>>>>>>> >>>>>> *>provided at
>>>>>>>>> >>>>>>http://www.pharmacy.**manchester.ac.uk/bryce/amber
>>>>>>>>> <http://manchester.ac.uk/bryce/amber> <http://www.pharmac
>>>>>>>>> >>>>>>y.manchester.ac.uk/bryce/amber
>>>>>>>>> <http://y.manchester.ac.uk/bryce/amber> >.
>>>>>>>>> >>>>>> The
>>>>>>>>> >>>>>> membrane *
>>>>>>>>> >>>>>> *>should be in melted phase (L_alpha phase) as reported in the
paper
>>>>>>>>> >>>>>> but
>>>>>>>>> >>>>>> I *
>>>>>>>>> >>>>>> *>am *
>>>>>>>>> >>>>>> *>getting the Gel phase at same temperature (303K) with in 10
ns
>>>>>>>>> >>>>>> simulation. *
>>>>>>>>> >>>>>> *>The reported area per lipid of DMPC at 303 K is ~61 square
>>>>>>>>> >>>>>>angstroms
>>>>>>>>> >>>>>> per *
>>>>>>>>> >>>>>> *>lipid but I am getting the area per lipid ~50 square
>>>>>>>>> angstroms per
>>>>>>>>> >>>>>> lipid *
>>>>>>>>> >>>>>> *>and membrane comes in gel phase. I would appreciate receiving
any
>>>>>>>>> >>>>>>*
>>>>>>>>> >>>>>> *>suggestion in this regard. *
>>>>>>>>> >>>>>> *> *
>>>>>>>>> >>>>>> *> *
>>>>>>>>> >>>>>> *> *
>>>>>>>>> >>>>>> *>Regards, *
>>>>>>>>> >>>>>> *>Subbarao Kanchi. *
>>>>>>>>> >>>>>> *>____________________________**___________________ *
>>>>>>>>> >>>>>> *>AMBER mailing list *
>>>>>>>>> >>>>>> *>AMBER.ambermd.org <http://AMBER.ambermd.org> *
>>>>>>>>> >>>>>>
>>>>>>>>>
>>>>>>*>http://lists.ambermd.org/**mailman/listinfo/amber<http://lists.ambe
>>>>>>>>> >>>>>>rmd.org/mailman/listinfo/amber
>>>>>>>>> <http://rmd.org/mailman/listinfo/amber> >*
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>>> >>>>>>
>>>>>>>> >>>>> ______________________________**_________________
>>>>>>> >>>> AMBER mailing list
>>>>>>> >>>> AMBER.ambermd.org
>>>>>>> >>>>
>>>>>>>
>>>>http://lists.ambermd.org/**mailman/listinfo/amber<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 Wed Oct 09 2013 - 05:30:03 PDT
