Re: [GaMD-discuss] Queries in the input

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

The force scale factor in gamd.log shows around 0.8-0.9 is fine. What I suggested to check is the k0P/k0D in the mdout file? It should be stay 1 as the collected potential for GaMD production is optimized.

Best,
Jinan

From: SRUTHI NAMBIAR <sru...@gm...>
Sent: Friday, March 26, 2021 1:14 AM
To: Wang, Jinan <ja...@ku...>
Cc: GaM...@li...
Subject: Re: [GaMD-discuss] Queries in the input

In the case of the equilibration stage, the total force weight and dihedral force weight did not reach to 1 exactly. It fluctuates between 0.89-0.99. Would that mean it is equilibrated well? How can I ensure that the gamd equilibration is correct. I have given the values of ntave, nteb and ntcmd as you have suggested. I am attaching my gamd.log and gamd.in<https://nam10.safelinks.protection.outlook.com/?url=http%3A%2F%2Fgamd.in%2F&data=04%7C01%7Cjawang%40ku.edu%7C6f1aaab52b2d4be6a1e808d8f01e755a%7C3c176536afe643f5b96636feabbe3c1a%7C0%7C0%7C637523360999216225%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=dvHSIswGrMnUPtVpaFNd8GkWFSv4GLhY2ntuEO6vOPA%3D&reserved=0> files here.

On Thu, Mar 18, 2021 at 11:45 PM Wang, Jinan <ja...@ku...<mailto:ja...@ku...>> wrote:
Yes, at each stage you need use the rst file from the previous stage. Same as for the two GaMD production.
You need to give different name of trajectory, mdout and gamd log file at different runs, otherwise the corresponding files are overwritten.

Best,
Jinan

From: SRUTHI NAMBIAR <sru...@gm...<mailto:sru...@gm...>>
Sent: Wednesday, March 17, 2021 10:18 PM
To: Wang, Jinan <ja...@ku...<mailto:ja...@ku...>>
Subject: Re: [GaMD-discuss] Queries in the input

Thanks a lot for the detailed explanation. Just a query on the restart files we should give at each stage. At each stage should I give the rst file of the previous stage? If the rst of gamd equilibration of gamd1.rst, should that be the -c for the gamd production? And what about the case where we divide the production as 2 parts?

Also incase I am restarting the simulation, will the nc file get overwritten starting from the point I am restarting?

Thanks for your help.

On Thu, 18 Mar 2021 at 5:46 AM, Wang, Jinan <ja...@ku...<mailto:ja...@ku...>> wrote:
Hi,

1.      It is better to set the value of ntave equal four times of atoms in the simulation system; ntcmdprep and ntebprep as two times of ntave; ntcmd as five times of ntave. The nteb should be long enough to get well equilibrated potential statistic. It could be tens of ns to hundreds of ns. One metric to determine is that both k0P and k0D should reach to 1 at the end of GaMD equilibration.Since your system has 200,000 atoms, it is better to use ntave=800,000, ntcmdprep=1,600,000, ntebprep=1,600,000. You can have a test of 60 ns for your nteb.

2.      It is better to separate the GaMD equilibration (like 64ns, keeping in mind that the nteb should be times of ntave) and  GaMD production (300ns) into two runs.

The GaMD equilibration input file should be like following:

&cntrl
  imin=0,irest=0,ntx=1,
  nstlim=36000000,dt=0.002,
  ntc=2,ntf=2, tol=0.000001,
  iwrap=1, ntb=1, cut=10.0,
  ntt=3, temp0=300.0, tempi=300.0,
  ntpr=1000, ntwx=1000, ntwr=500,
  ntxo=1, ioutfm=1, ig=-1, ntwprt=27735,
  igamd=3, iE=1, irest_gamd=0,
  ntcmd=4000000, nteb = 32000000, ntave = 800000,

 ntcmdprep = 1600000, ntebprep = 1600000,
  sigma0P = 6.0, sigma0D = 6.0,
 &end

300 ns GaMD production in one simulation, the input file:

 &cntrl
  imin=0,irest=0,ntx=1,
  nstlim=150000000,dt=0.002,
  ntc=2,ntf=2, tol=0.000001,
  iwrap=1, ntb=1, cut=10.0,
  ntt=3, temp0=300.0, tempi=300.0,
  ntpr=1000, ntwx=1000, ntwr=500,
  ntxo=1, ioutfm=1, ig=-1, ntwprt=27735,
  igamd=3, iE=1, irest_gamd=1,
  ntcmd=0, nteb = 0, ntave = 800000,

 ntcmdprep = 0, ntebprep = 0,
  sigma0P = 6.0, sigma0D = 6.0,
 &end

3.      For splitting the 300 ns GaMD production into two 150 ns simulations. The first one just change the above nstlim=150000000 into nstlim=75000000. And the second one using the following input file

&cntrl
  imin=0,irest=1,ntx=5,
  nstlim=75000000,dt=0.002,
  ntc=2,ntf=2, tol=0.000001,
  iwrap=1, ntb=1, cut=10.0,
  ntt=3, temp0=300.0, tempi=300.0,
  ntpr=1000, ntwx=1000, ntwr=500,
  ntxo=1, ioutfm=1, ig=-1, ntwprt=27735,
  igamd=3, iE=1, irest_gamd=1,
  ntcmd=0, nteb = 0, ntave = 800000,

 ntcmdprep = 0, ntebprep = 0,
  sigma0P = 6.0, sigma0D = 6.0,
 &end

Hoping the above information will help.
Best,
Jinan
From: SRUTHI NAMBIAR <sru...@gm...<mailto:sru...@gm...>>
Sent: Wednesday, March 17, 2021 4:51 AM
To: GaM...@li...<mailto:GaM...@li...>
Subject: [GaMD-discuss] Queries in the input

Dear all,
I have recently started to do GaMD simulations in AMBER18 for a 200,000 atoms Protein-RNA system. Since I am a beginner, I am getting confused in the input creation and some trivial doubt.

1. If I have to run  a simulation with ~300 ns (GaMD production) ,  60 ns (GaMD equilibration) + 10 ns (pre-equilibration classical MD), in which the boost potential has been updated every 1.5 ns is it enough to give the following input parameters:
nstlim = 150000000 dt = 0.002
ntcmd = 5000000 nteb =   30000000 ntave = 50000
ntcmdprep = 200000, ntebprep = 800000

2. Incase I am running the whole 300ns in a single run (no walltime limit), would the following script be sufficient?
 &cntrl
  imin=0,irest=0,ntx=1,
  nstlim=150000000,dt=0.002,
  ntc=2,ntf=2, tol=0.000001,
  iwrap=1, ntb=1, cut=10.0,
  ntt=3, temp0=300.0, tempi=300.0,
  ntpr=1000, ntwx=1000, ntwr=500,
  ntxo=1, ioutfm=1, ig=-1, ntwprt=27735,
  igamd=3, iE=1, irest_gamd=0,
  ntcmd=5000000, nteb = 30000000, ntave = 50000,
  ntcmdprep = 200000, ntebprep = 800000
  sigma0P = 6.0, sigma0D = 6.0,
 &end

3. Incase I am doing my simulation in 2 stages, JOB1 AND JOB2 150ns + 150 ns, I am a bit doubtful in how I should give the command line so that gamd-restart.dat will be read.

JOB 1 :
  imin=0,irest=0,ntx=1,
  nstlim=150000000,dt=0.002,
  ntc=2,ntf=2, tol=0.000001,
  iwrap=1, ntb=1, cut=10.0,
  ntt=3, temp0=300.0, tempi=300.0,
  ntpr=1000, ntwx=1000, ntwr=500,
  ntxo=1, ioutfm=1, ig=-1, ntwprt=27735,
  igamd=3, iE=1, irest_gamd=0,
  ntcmd=5000000, nteb = 30000000, ntave = 50000,
  ntcmdprep = 200000, ntebprep = 800000
  sigma0P = 6.0, sigma0D = 6.0,
 &end

pmemd.cuda -O -i gamd.in<https://nam10.safelinks.protection.outlook.com/?url=http%3A%2F%2Fgamd.in%2F&data=04%7C01%7Cjawang%40ku.edu%7C6f1aaab52b2d4be6a1e808d8f01e755a%7C3c176536afe643f5b96636feabbe3c1a%7C0%7C0%7C637523360999216225%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=dvHSIswGrMnUPtVpaFNd8GkWFSv4GLhY2ntuEO6vOPA%3D&reserved=0> -o gamd.out -p prot.prmtop -c prod.rst7 -r gamd1.rst7 -x gamd.nc<https://nam10.safelinks.protection.outlook.com/?url=http%3A%2F%2Fgamd.nc%2F&data=04%7C01%7Cjawang%40ku.edu%7C6f1aaab52b2d4be6a1e808d8f01e755a%7C3c176536afe643f5b96636feabbe3c1a%7C0%7C0%7C637523360999226222%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=Mau4vzRf3%2FicaChKVOR%2FgxvSRNIZybakPQnZPtjyzm8%3D&reserved=0>

Job 2:
  imin=0,irest=1,ntx=1,
  nstlim=150000000,dt=0.002,
  ntc=2,ntf=2, tol=0.000001,
  iwrap=1, ntb=1, cut=10.0,
  ntt=3, temp0=300.0, tempi=300.0,
  ntpr=1000, ntwx=1000, ntwr=500,
  ntxo=1, ioutfm=1, ig=-1, ntwprt=27735,
  igamd=3, iE=1, irest_gamd=1,
  ntcmd=0, nteb = 0, ntave = 50000,
  ntcmdprep = 200000, ntebprep = 800000
  sigma0P = 6.0, sigma0D = 6.0,
 &end

  pmemd.cuda -O -i gamd.in<https://nam10.safelinks.protection.outlook.com/?url=http%3A%2F%2Fgamd.in%2F&data=04%7C01%7Cjawang%40ku.edu%7C6f1aaab52b2d4be6a1e808d8f01e755a%7C3c176536afe643f5b96636feabbe3c1a%7C0%7C0%7C637523360999236217%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=LJupxxaQKGd%2FvqE3WXymulf264FhflISr2HY3oUxDjs%3D&reserved=0> -o gamd.out -p prot.prmtop -c gamd1.rst7 -r gamd2.rst7 -x gamd.nc<https://nam10.safelinks.protection.outlook.com/?url=http%3A%2F%2Fgamd.nc%2F&data=04%7C01%7Cjawang%40ku.edu%7C6f1aaab52b2d4be6a1e808d8f01e755a%7C3c176536afe643f5b96636feabbe3c1a%7C0%7C0%7C637523360999246213%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=KOcOL08RwLRdzDVyfn338U0xrTsA%2FszoQ8gjfXt6%2B5Y%3D&reserved=0>

Would the above scripts and command lines be sufficient? Kindly answer and help.