Oh I just saw your goal at the end of the email - in my opinion, tgtmd is
very unlikely to give you a reasonable transition state. This is quite
challenging in general and you might want to do a literature search to see
how it has been approached by others. Not something you're likely be be
able to finish on your tight deadline unfortunately.
On Tue, Sep 5, 2023, 1:10 PM Malcolm Lim <mjyl3.cam.ac.uk> wrote:
> Hi Adrian and Carlos,
>
> Many thanks for your replies. I've followed the input files of a few other
> users which have also performed TMD in the past 20 years in the archives
> and have scrapped together the input file, hence explaining its utter
> incoherence.
>
> I've made the following modifications to the input file based on your
> recommendations as below, along with some questions that I have.
>
> .Carlos Simmerling:
> 1. I've made ntb and ntp consistent at 0, as my understanding is that
> aligning to boundary conditions and any pressure is not essential for this
> purpose.
> 2. I've also changed the relevant parameter to tgtfitmask.
> 3. Understood that tgtmdfrc is really small - had referenced that from a
> prior post (which had a different purpose: trying to fit a ligand into the
> protein). I've adjusted mine to 0.5 - not too sure what is the ideal value,
> but I would presume that would need some trial and error.
>
> Q1. Is there a way to reliably minimize the number of steps (e.g. in a
> shorter time frame or adjusting nstlim/dt/ntpr/ntwx values) and therefore
> reduce the amount of time run for each trial run, so I can reduce the
> overall amount of time for optimizing the necessary conditions for an ideal
> run?
>
> .Adrian Roitberg:
> 1. I've removed tempi and temp0 values as I would very much prefer
> changing the structure as opposed to heating (got a little confused here as
> I saw another post doing TMD but including the temperature parameters, and
> I thus thought it would be essential for such an action - thanks for the
> clarification).
>
> 2. Apologies for my ignorance regarding the dt value: I had initially
> increased it with the idea of minimizing the time required for trial runs
> to test the parameters but saw that the maximum that should be used is
> indeed 0.002. Thank you for that.
>
> 3. I've changed the input to ntr=0 and have removed all restraint-related
> parameters. I would presume in my case of doing a TMD, there would be no
> need for a restraint component other than having the target (ref) structure
> as the destination for the structural coordinates.
>
> Q1. May I ask which tutorials would you find the most relevant for such a
> purpose for doing TMD? I understand that this is an additional feature atop
> traditional MD simulations which does not possess a dedicated tutorial for,
> which results in me trying to scrap together what I've learnt from various
> tutorials and trying to implement them together as a pice.
>
> For reference, here is the current input file:
> TMD
> &cntrl
> imin=0,
> ntx=1,
> irest=0,
> nstlim=10000,
> dt=0.002,
> ntf=2,
> ntc=2,
> ntpr=100,
> ntwx=100,
> cut=16.0,
> ntb=0,
> ntp=0,
> ntt=3,
> gamma_ln=2.0,
> itgtmd=1,
> tgtmdfrc=0.5,
> ntr=0,
> tgtfitmask=":1-279",
> tgtrmsmask=":1-279"
> /
> &wt type='TGTRMSD', istep1=0.0, istep2=10000, value1=3.663, value2=0.0
> /
> &wt type='END'
> /
>
> Which eventually gives me an error of "NATOM mismatch in constraint coord
> and topology files", leading to another question: Does this mean that the
> number of atoms present in the ref inpcrd and input inpcrd/prmtop must be
> the same? (as they were solvated separately and had solvent boxes of
> different sizes).
>
> Apologies for the rush to optimize this process in a short time, I'm
> currently working on a tight deadline trying to find the transition state
> structure between the said two states and thought of TMD as a possible
> method to do so, thus my interest in AMBER. If there are any other better
> alternatives to this, please do let me know as well.
>
> Once again, thank you very much for the input and taking the time to read
> this, and on top of all, your patience in dealing with my ignorance.
>
> Best wishes,
> Malcolm
>
> -----Original Message-----
> From: Adrian Roitberg via AMBER <amber.ambermd.org>
> Sent: Tuesday, September 5, 2023 4:46 PM
> To: amber.ambermd.org
> Subject: Re: [AMBER] Clarification on attempted TMD run
>
> Also, you need to run some of the tutorials before you even try TGTMD
>
> Plenty of problem with your input, such as you trying changing the
> temperature (heating) and changing structure at the same time, using a
> dt=0.2 instead of 0.002, and as far as I can tell, restraining your
> structure with ntr AND attempting to change it at the same time.
>
> Adrian
>
>
> On 9/5/23 11:07 AM, Carlos Simmerling via AMBER wrote:
> > [External Email]
> >
> > You have a typo, check the manual section for tgtmd. It should be
> > tgtfitmask not tgtmdmask.
> > Also make sure you have run stable md on the system using the same
> > inputs (other than tgtmd). Here your ntb and ntp seem inconsistent.
> > Your force constant for tgtmd is small. Keep that in mind if you
> > aren't able to achieve the transition.
> >
> > On Tue, Sep 5, 2023, 10:35 AM Malcolm Lim via AMBER
> > <amber.ambermd.org>
> > wrote:
> >
> >> Dear Sir/Ma'am,
> >>
> >> I am new to AMBER and am trying to use AMBER22 for a Targeted MD
> >> (TMD) simulation, with my objective being trying to simulate the
> >> transition between 2 conformational states of a specific protein. I
> >> am starting with the .pdb files of both conformational states.
> >>
> >> I've read through the manual and have prepared the pdb files via
> >> tleap, and have used the following script to prepare both structures
> >> in tleap (e.g. for state 'c' to state 'm'):
> >> #tleap_c.in
> >> source leaprc.protein.ff19SB
> >> source leaprc.water.opc
> >> #loadamberparams frcmod.ions1lm_126_hfe_opc c=loadpdb
> >> cState.amber.pdb SaveAmberParm c c_gas.prmtop c_gas.inpcrd addIons c
> >> Na+ 13 solvateOct c OPCBOX 10.0 addIonsRand c Na+ 35 Cl- 35
> >> SaveAmberParm c c_ion.prmtop c_ion.inpcrd
> >>
> >> Using the prmtop and inpcrd files, I've attempted to run the TMD on
> >> sander via the following command:
> >>
> >> sander -O -i tmd.in -o tmd.out -p c_ion.prmtop -c c_ion.inpcrd -r
> >> tmd.ncrst -ref m_ion.inpcrd -x tmd.nc
> >>
> >> where tmd.in contains the following:
> >>
> >> TMD
> >> &cntrl
> >> imin=0,
> >> ntx=1,
> >> irest=0,
> >> nstlim=10000,
> >> dt=0.2,
> >> ntf=2,
> >> ntc=2,
> >> tempi=0.0,
> >> temp0=300.0,
> >> ntpr=100,
> >> ntwx=100,
> >> cut=8.0,
> >> ntb=2,
> >> ntp=0,
> >> ntt=3,
> >> gamma_ln=2.0,
> >> itgtmd=1,
> >> tgtmdfrc=0.005,
> >> ntr=1,
> >> restraint_wt=10.0,
> >> restraintmask=":1-279",
> >> tmdmask=":1-279",
> >> tgtrmsmask=":1-279"
> >> /
> >> &wt type='TGTRMSD', istep1=0.0, istep2=10000, value1=3.663,
> >> value2=0.0 / &wt type='END'
> >> /
> >>
> >> However, I've run into an error: "error in reading namelist cntrl",
> >> which indicates that perhaps the syntax of the .in file is amiss,
> >> though I'm uncertain of what is wrong. I just have 2 questions for the
> above:
> >>
> >>
> >> 1. Could I check if there is anything wrong with the .in file
> >> that is causing this issue? Or perhaps something wrong with the
> >> parameters in the command line and/or in the preparation of the
> necessary prmtop/inpcrd files?
> >> 2. Could I also ask if the parameters in the .in file are
> >> appropriate for the purpose of this TMD run? Or am I missing out
> >> other factors that should be considered when executing such a
> simulation of this nature?
> >>
> >> Thank you very much and looking forward to learning more about AMBER!
> >>
> >> Best wishes,
> >> Malcolm
> >> _______________________________________________
> >> AMBER mailing list
> >> AMBER.ambermd.org
> >> http://list/
> >> s.ambermd.org%2Fmailman%2Flistinfo%2Famber&data=05%7C01%7Cmjyl3%40uni
> >> versityofcambridgecloud.onmicrosoft.com%7C2b4f6564f4da43a685b208dbae2
> >> 7b848%7C49a50445bdfa4b79ade3547b4f3986e9%7C1%7C0%7C638295257842537225
> >> %7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6
> >> Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=H0%2B3XGyNwF78W2Mhfh5mNB
> >> Yt9wOcs9L3%2BtyikaOpUp0%3D&reserved=0
> >>
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists/
> > .ambermd.org%2Fmailman%2Flistinfo%2Famber&data=05%7C01%7Cmjyl3%40unive
> > rsityofcambridgecloud.onmicrosoft.com%7C2b4f6564f4da43a685b208dbae27b8
> > 48%7C49a50445bdfa4b79ade3547b4f3986e9%7C1%7C0%7C638295257842537225%7CU
> > nknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1ha
> > WwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=H0%2B3XGyNwF78W2Mhfh5mNBYt9wOc
> > s9L3%2BtyikaOpUp0%3D&reserved=0
>
> --
> Dr. Adrian E. Roitberg
> V.T. and Louise Jackson Professor in Chemistry Department of Chemistry
> University of Florida roitberg.ufl.edu
> 352-392-6972
>
>
> _______________________________________________
> 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 Sep 05 2023 - 11:30:02 PDT