I am sorry for my late reply.
I can run newton.x with the generated xyz file and amoebabio09.prm. Although it is very slow, it is in progress.
The output is like this:
TN Iter F Value G RMS F Move X Move CG Iter Solve FG Call
0 -62163.7371 22.2815 1
HESSIAN -- 2903358 Elements 0.31 % Off-Diag Hessian Storage
1 -62170.9424 22.2708 7.2053 0.0030 2 NegCurve 6
HESSIAN -- 2903682 Elements 0.31 % Off-Diag Hessian Storage
2 -62800.0336 20.7270 629.0913 0.0156 5 NegCurve 12
HESSIAN -- 2903763 Elements 0.31 % Off-Diag Hessian Storage
BTW, tinker 6 in my server was compiled with libfftw3_omp.a and openmp enabled (ifort -c -O3 -xHost -no-ipo -no-prec-div -assume cc_omp -openmp XXX.f).
I set OMP_NUM_THREADS=8 and added "OPENMP-THREADS 8" in the key file.
But newton.x occupies only one CPU core. I can hardly find the %CPU a little bit higher than 100.
Is there anything else I have to do to enable the parallel run?
All the best,
Ye
2013-03-13
From： Amber Amoeba
Date： 2013-03-12 02:18:44
To： AMBER Mailing List
CC：
Subject： Re: [AMBER] bug report: tinker_to_amberfailedforsystemscontainingcounter ions
Can you optimize the structure in TINKER before converting to amber? Use
parallel openmp-thread will help speed up a bit.
It seems that all terms are reasonable except for the angle. I would double
check the structure and connectivity. Or start with a protein PDB/structure
that is already optimized.
On Fri, Mar 8, 2013 at 9:35 AM, Ye MEI <ymei.itcs.ecnu.edu.cn> wrote:
> Hi, Prof. Ren,
>
> I still cannot get it work.
> I tried minimization with the attached input file. Other relevant files
> can be downloaded from
> http://itcs.ecnu.edu.cn/ymei/AmoebaTest.zip
>
> The output energy is weird with abnormally large bending energy
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 1 1.5700E+11 5.3046E+05 1.9333E+06 O 26054
> BOND = 1562.3590 ANGLE = ************* DIHED =
> 395.9413
> VDWAALS = 28514.7995 EEL = -72710.5133 HBOND =
> 0.0000
> 1-4 VDW = 2171.7723 1-4 EEL = 0.0000 RESTRAINT =
> 0.0000
> EPOLAR = -27008.8684
> Dipole convergence: rms = 0.985E-02 iters = 10.00
> out of loop, iter,rms = 10 0.9847E-02
> e_rec_perm,e_dir_perm,e_adj_perm,e_self_perm =
> 5026.8137 -55712.3506 275540.1023 -297570.4186
> e_rec_ind,e_dir_ind,e_adj_ind,e_self_ind =
> -470.1890 -21441.8769 -4358.6416 -741.0962
> e_dir_vdw,e_adj_vdw,e_rec_vdw =
> 28730.6905 2171.7723 -212.2151
> nonbond vir = -49611.408 4194.7509 -2144.7599
> nonbond vir = 4194.7509 -51124.669 2564.9863
> nonbond vir = -2144.7599 2564.9863 -48412.859
> valence energies: bond,ureyb,angle 0.15623844E+04 0.15699537E+12
> 0.20018563E+04
> valence energies: trangle,opbend,tor 796.3284 179.4080
> 401.4212
> valence energies: pitors,strbend,tortor 16.7169 -21.4293
> -22.1968
> valence energies: strtor 0.0000
> valence vir = ************** 214915029.1693 83415569.4066
> valence vir = 214915029.1693 ************** **************
> valence vir = 83415569.4066 ************** **************
> out of loop, iter,rms = 10 0.9847E-02
> e_rec_perm,e_dir_perm,e_adj_perm,e_self_perm =
> 5027.1092 -55717.6459 275538.6938 -297570.4186
> e_rec_ind,e_dir_ind,e_adj_ind,e_self_ind =
> -470.1654 -21444.7449 -4359.2458 -741.1707
> e_dir_vdw,e_adj_vdw,e_rec_vdw =
> 28735.1026 2171.7723 -212.2151
>
> Is it due that the structure was generated by LEaP without any
> optimization?
>
>
> Ye
> 2013-03-08
>
>
> From： Amber Amoeba
> Date： 2013-03-08 00:05:44
> To： AMBER Mailing List
> CC：
> Subject： Re: [AMBER] bug report: tinker_to_amber
> failedforsystemscontainingcounter ions
>
> The "atomic charge" is not used. Search for the
> "AMOEBA_LOCAL_FRAME_MULTIPOLES_LIST". You should see the charges/dipoles
> etc there.
> I can't tell what's wrong with the lines you posted. Did you try to run
> energy minimization (you may turn off polarization first and then back on).
> On Thu, Mar 7, 2013 at 4:31 AM, Ye MEI <ymei.itcs.ecnu.edu.cn> wrote:
> > Hi, Prof. Ren,
> >
> > I cannot compile tinker4.3 in my x64 machine. It failed with error
> message
> > undefined reference to `chksocket_'
> > Including server.f in the source folder did not solve the problem.
> > So I followed the second way. All the steps seemed fine.
> > Tinker_to_amber gave the following output:
> > frcfieldfile = /home/ymei/tinker6/params/amoebabio09.prm
> > found a-axis = 68.5238760000000
> > found b-axis = 71.8084030000000
> > found c-axis = 68.2765680000000
> > found alpha = 90.0000000000000
> > found beta = 90.0000000000000
> > found gamma = 90.0000000000000
> > num_xyz_atoms = 26140
> > numres = 8048
> > num_atoms = 26140
> > vdw: num_vdw_types = 19
> > num_bonds = 18273
> > num_urey_bradley = 7893
> > num_tot_angles = 12421
> > num regular, trigonal angles = 10999 1422
> > num_out_of_plane_bends = 1422
> > numtor = 6619
> > num1,num_torsions = 5747 8964
> > no success finding Stretch-Torsions
> > num_pitorsions = 268
> > num_stretch_bends = 3955
> > num_tor_tors = 147
> > num tortor tables = 8
> > num_molecules = 7896
> > num_multipoles = 26140
> > number of frame def list = 60199
> > num_12 = 18273
> > num_13 = 12421
> > num_14 = 6529
> > num_15 = 7403
> >
> > I used a cuboid box this time and I recompiled the code with maxatm in
> > sizes.i changed to 50000.
> > But the generated prmtop file was strange.
> > I paste some of the lines here
> >
> > %FLAG POINTERS
> > %FORMAT(10I8)
> > 26140 1 1 1 1 1 1 1 0
> > 0
> > 44616 8048 1 1 1 1 1 1 1
> > 1
> > 0 0 0 0 0 0 0 1 0
> > 0
> > 0 0
> >
> > and all the atomic charges were zero.
> >
> >
> > Best regards,
> >
> > Ye
> > 2013-03-07
> >
> >
> > From： Amber Amoeba
> > Date： 2013-03-07 14:54:50
> > To： AMBER Mailing List
> > CC：
> > Subject： Re: [AMBER] bug report: tinker_to_amber failed
> > forsystemscontainingcounter ions
> >
> > Ye,
> > You have a couple of options.
> > 1. You can use the older tinker v4.3 (found here
> > http://dasher.wustl.edu/tinker/). If you want to use amoebabio09.prm
> with
> > this version, you need to back-convert it using amoeba_conv.pl (download
> > from http://water.bme.utexas.edu/wiki/index.php/Software:amber). I would
> > recommend using the "older" version amoebapro04.prm (
> > http://water.bme.utexas.edu/wiki/index.php/Software:amoeba_bio).
> > 2.If you want to use the latest tinker like you did, please replace
> > analyze.f with this one:
> > http://biomol.bme.utexas.edu/~pren/tinker_to_amber/.
> > This fix the ion issue and connectivities that were missing. You still
> have
> > to change the "None" in the analout to "Z-then-X" manually for ions.
> Please
> > do not change the tinker_to_amber source code. Ignore the warning about
> the
> > torsion-torsion spline. It is due to small mismatches at the boundary
> (e.g.
> > values 180 and -180 should be exactly the same but actually differ a tiny
> > bit).
> > 3. No matter you choose 1 or 2, if you are using an octahedron box, you
> > need to make some manual changes to the inpcrd and prmtop files produced
> by
> > the tinker_to_amber. The format is explained in amber manual. The
> specific
> > changes are explained here
> > http://water.bme.utexas.edu/wiki/index.php/Software:amber
> > Please let me know if you have problems. These changes will be
> incorporated
> > into future tinker/amber.
> > Thanks
> > Pengyu
> > On Wed, Mar 6, 2013 at 8:23 PM, Ye MEI <ymei.itcs.ecnu.edu.cn> wrote:
> > > I just put the pdb file into the zipped folder. Please download it
> again.
> > > http://itcs.ecnu.edu.cn/ymei/tinker2amber.zip
> > >
> > > I forgot to mention that I had moved the atomic name of Cl- one column
> to
> > > the right before running pdbxyz.
> > > Otherwise the code cannot recognize it.
> > >
> > > Ye
> > > 2013-03-07
> > >
> > >
> > > From： Amber Amoeba
> > > Date： 2013-03-07 02:29:16
> > > To： AMBER Mailing List
> > > CC：
> > > Subject： Re: [AMBER] bug report: tinker_to_amber failed for
> > > systemscontainingcounter ions
> > >
> > > Ye,
> > > Can you also provide the matching pdb file? There is some format
> changes
> > in
> > > the recent tinker. I will get it fixed.
> > > On Wed, Mar 6, 2013 at 2:19 AM, Ye MEI <ymei.itcs.ecnu.edu.cn> wrote:
> > > > Hi, Prof. Ren,
> > > >
> > > > All the relevant files can be found here:
> > > > http://itcs.ecnu.edu.cn/ymei/tinker2amber.zip
> > > >
> > > > The protein myoglobin was solvated in a truncated octahedron water
> box
> > > > with two
> > > > chloride ions. The whole system was generated by LEaP module in
> > > AmberTools
> > > > 12.
> > > >
> > > > Then I split this pdb file into two files. One only contained the
> > > protein.
> > > > The other file
> > > > contained all the atoms. Keeping protein atoms in the second file is
> to
> > > > maintain the
> > > > correct sequence number for atoms and their connectivities.
> > > >
> > > > I used pdbxyz command to convert two pdb files to xyz files with
> > > > amoebabio09.prm.
> > > > PDB file for the protein was OK. But ions and water molecules need
> some
> > > > modifications.
> > > > I used the program fix_wat_xyz_amoeba_type.f released with AMBER 12
> to
> > > > change the
> > > > atom types to 402 and 403 for oxygen and hydrogen atoms in water
> > > > molecules, and 412 for
> > > > the chloride ions.
> > > >
> > > > Then I removed protein atoms in the second xyz file, and combined two
> > xyz
> > > > files together.
> > > > I also changed the number of atoms in the first line of the combined
> > xyz
> > > > file.
> > > > The ouput of "analyze.x peptide_wat PC >peptide_wat_analout" also
> > seemed
> > > > fine.
> > > >
> > > > Then I typed 'tinker_to_amber -name peptide_wat -title "amoeba
> > myoblogin"
> > > > '. It gave warning message:
> > > > ISPLPE -- Warning, Input Values to Spline are Not Periodic
> > > > and error message
> > > > wrong number of fields in multipole line: numf = 4
> > > >
> > > > I have not figured out how to deal with the warning message. I will
> > check
> > > > the manuals.
> > > > The error message was caused by the chloride ions. In line 2459-2476
> of
> > > > tinker_to_amber.f90,
> > > > it only handles lines with 6 or 7 fields, and gives error message for
> > > > other lines. So I changed this
> > > > part and you can find the code in the zipped file. This time it
> > finished
> > > > with only warning messages:
> > > > ISPLPE -- Warning, Input Values to Spline are Not Periodic
> > > > no success finding Number of 1-2 Pairs!
> > > > no success finding Number of 1-3 Pairs!
> > > > no success finding Number of 1-4 Pairs!
> > > > no success finding Number of 1-5 Pairs!
> > > >
> > > > I guess there may be something wrong with the key file.
> > > > The prmtop file was not correctly generated.
> > > >
> > > > So now I am wondering
> > > > 1. Is the new tinker_to_amber.f90 code correct?
> > > > 2. Should I add more control options into the key file?
> > > >
> > > > I am using tinker-6.1.01 compiled with intel compiler in an x64
> > machine.
> > > > I can run md simulations with command dynamic.x.
> > > >
> > > > All the best,
> > > >
> > > > Ye
> > > > 2013-03-06
> > > >
> > > >
> > > >
> > > > From： Amber Amoeba
> > > > Date： 2013-03-06 04:34:19
> > > > To： ymei; amber
> > > > CC：
> > > > Subject： [AMBER] bug report: tinker_to_amber failed for systems
> > > > containingcounter ions
> > > >
> > > > Hi Ye,
> > > > Can you tell me which version of tinker (analyze) are you using?
> > > > How did you add the metal ions into tinker xyz and PDB files?
> > > > After you prepare the tinker files, before you run tinker_to_amber,
> > were
> > > > you able to run tinker analyze or minimize and get an energy?
> > > > Thanks
> > > > Pengyu Ren
> > > > UT Austin
> > > > _______________________________________________
> > > > 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
> _______________________________________________
> 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 Mar 13 2013 - 08:00:03 PDT
