On Fri, Apr 6, 2018 at 10:26 PM, Alexander Alexander <alexanderwie...@gmail.com> wrote: > > Dear Christoph, > > On Thursday, April 5, 2018 at 6:46:13 AM UTC-4, Christoph Junghans wrote: >> >> On Wed, Apr 4, 2018 at 5:52 PM, Alexander Alexander >> <alexand...@gmail.com> wrote: >> > Thanks Christoph. >> > >> > Finally my IBI (both bonded and nonbonded) is running know, but there is >> > a >> > problem and that is the very high temperature (around 11842 K) which I >> > get >> > by gmx energy over the ene.edr of each step while I use ref_t = 298.15 K >> > in >> > my .mdp file using sd integrator. More or less I am using the grompp.mdp >> > file used in the tutorial examples. >> > Here is a part of it: >> > integrator = sd >> > tcoupl = no >> > tc-grps = System >> > tau_t = 1.0 >> > ref_t = 298.15 >> > Pcoupl = no >> > pcoupltype = isotropic >> > tau_p = 0.5 >> > compressibility = 4.5e-5 >> > ref_p = 1.0 >> > gen_vel = yes >> > gen_temp = 298.15 >> > gen_seed = -1 >> > >> > I do not know what exactly is going on. >> Just a bunch on generic tips: >> Run longer, try a smaller timestep, watch the simulation in VMD to see >> if something looks strange. > > > The temperature is now fine around 298.15 k, to reach that although I had to > use dt = 0.00002 but the key point was tau_t which I changed it to 0.02 from > 1.0 to have the temperature correct finally. The dt seems extremely small and the tau_t uncommonly large. It seems to me, you just made the heat-up process very slow with such a small dt, you should really check setup again.
> > But the pressure is too high around 1143.92 in each step, which I found it > that is normal when I compared with your spce-ibi example which has 6000 or > 8000 of pressure in each step, am I right, please? Yeah, that is common, you could to pressure correcting later down the road. > > One more issue: > > In my IBI (bonded and nonbonded), I get errors like > ""inverse/dist_boltzmann_invert.pl --type dihedral --kbT 2.4789711565 --min > 1e-10 bond-LA-LB.dist.tgt bond-LA-LB.pot.new.raw.NK3 " for the bonds and > angles. I can solve it by decreasing the step to <step>0.00001</step> for > them, which then slows down everything and also the improvement in > potentials and ... . What could be the reason for such an error? Sorry what is the error message? And again you can set step to 0.00001, but that seems uncommonly small and you will have to collect 100 times more statistics (= 100 longer simulation) if you make the bins size 100 times smaller. Christoph > > > Thank you. > Best regards, > Alex >> >> >> > >> > >> > On Wednesday, April 4, 2018 at 9:04:27 AM UTC-4, Christoph Junghans >> > wrote: >> >> >> >> On Tue, Apr 3, 2018 at 11:05 PM, Alexander Alexander >> >> <alexand...@gmail.com> wrote: >> >> > Thank you Christoph. >> >> > >> >> > The gmx_mpi mdrun crashed in the step_001, with the below error; >> >> > >> >> > Tabulated interaction of type 'Tab. Bonds' or 'Tab. Bonds NC' with >> >> > index >> >> > 0 >> >> > cannot be used because no table file whose name matched '_b0.xvg' was >> >> > passed >> >> > via the gmx mdrun -tableb command-line option. >> >> Yeah, there was a bug in gromacs-5.1 and since then you need to >> >> explicitly use "-tableb" in mdrun call. >> >> Usually VOTCA adds that automatically except if you have done it >> >> manually in settings xml file. >> >> It seems, you mis-numbered the tables in your topol.top: >> >> [ bonds ] >> >> 1 2 8 0 1 ; 1:bond-LG-H:1 >> >> 2 3 8 1 1 ; 1:bond-H-L:1 >> >> 3 4 8 1 1 ; 1:bond-L-RG:1 >> >> says you need a table_b0.xvg and table_b1.xvg (4th column is the table >> >> number - you using 0 and 1), but then you referring to table_b1.xvg, >> >> table_b2.xvg and table_b3.xvg in your settings.xml, so something is >> >> wrong here. >> >> >> >> (See the hexane/ibi_all for tutorial for details) >> > >> > I see now, I was not familiar with the [ bond ] (type 8) in tabulated >> > potential. >> >> >> >> >> >> > >> >> > But, it worked when I added "-tableb table_b0.xvg table_b1.xvg >> >> > table_b2.xvg >> >> > table_a0.xvg table_a1.xvg" to the gmx_mpi mdrun, I was surprised >> >> > because >> >> > it >> >> > had thought the -tableb is kind of default when one intend to use >> >> > tabulated >> >> > potential calming in case.mdp file. (gromacs2018.1). >> >> As you aren't referring to table_b0.xvg anywhere in your settings.xml >> >> file, I am not sure where this table is coming from. >> >> >> >> > >> >> > More questions; >> >> > >> >> > Similar to the normal atomistic simulation, I was wondering if the >> >> > minimization and equalization are needed for IBI (bonded and >> >> > nonbonded) >> >> > prior to the sd simulation in each step? >> >> VOTCA can do both, a minimization can be done before the main >> >> simulation (see the spce/simplex/simple example). >> >> For equalization, just set the <equi_time> in the xml file to >> >> something bigger than 0 and VOTCA will skip these parts of the >> >> simulation (similar to gmx's -b option) >> >> But if you need one of the two options is system dependent. >> >> >> >> > >> >> > By csg_map one can map the atomistic.trr to CG.xtc, what could be the >> >> > usage >> >> > of CG.txc in IBI (bonded and nonbonded) then? >> >> E.g. if you want to calculate some other properties of the CG system >> >> with gmx's tools, like diffusion constant. >> >> >> >> > >> >> > Should the conf.gro used in IBI be correspond to the conf.gro used in >> >> > the >> >> > all atom simulation? To be clear, if the CG.gro used in IBI must be >> >> > exactly >> >> > the one obtained by mapping the AA.gro and can not be bigger than >> >> > that? >> >> > Can >> >> > we make the CG.gro obtained by mapping twice to use it in IBI? >> >> Actually it is very common to scale up the system for IBI to get more >> >> statistics. >> >> And you don't have to use exactly the same starting configuration, but >> >> you should use one which is equilibrated (which is the case for mapped >> >> ones.) >> >> >> > You mean we can generate a new bigger CG-conf.gro using for example gmx >> > insert-molecules, equilibrate it before any IBI. The equilibration >> > should be >> > performed in CG level? If so, then who? because we still do not have any >> > good CG parameter for it to use them for equilibrating the scaled >> > CG-conf.gro. >> > Probably, we have to scale the system in atomistic level and equilibrate >> > it >> > still in atomistic level and then mappe the scaled-equilibrated.gro to >> > have >> > a CG-scaled-equilibrated.gro. >> You are right as the potential changes, "equilibrated" might have been >> the wrong word. >> I meant an initial condition without atoms overlaps like the mapped one. >> To build a bigger initial condition you can use e.g. "gmx genconf -nbox". >> >> Christoph >> > >> > Best regards, >> > Alex >> >> >> >> > >> >> > In using gmx rdf, we usually compare a selection (-sel) like A with a >> >> > reference (-ref) like B, and this means rdf_A_B.xvg might differ with >> >> > the >> >> > situation where we switch the reference and selection to have >> >> > rdf_B_A.xvg. >> >> > So, should we also consider this in for example csg_stat in >> >> > setting,xml >> >> > or >> >> > map.xml? in other words, should we define both <nonbonded> A B >> >> > </nonbonded> >> >> > and <nonbonded> B A </nonbonded>? >> >> Honestly, there is only one interaction between A and B, so why would >> >> you calculate two RDFs? >> >> Also the way csg_stat calculates RDFs is symmetric. >> >> >> >> Christoph >> >> >> >> > >> >> > Thanks for your time in advance. >> >> > >> >> > Best regards, >> >> > Alex >> >> > >> >> > On Tuesday, April 3, 2018 at 5:29:00 AM UTC-4, Christoph Junghans >> >> > wrote: >> >> >> >> >> >> >> >> >> >> >> >> On Tue, Apr 3, 2018, 09:31 Alexander Alexander >> >> >> <alexand...@gmail.com> >> >> >> wrote: >> >> >>> >> >> >>> Dear all, >> >> >>> >> >> >>> I am iterating over the both bonded and nonbonded interactions in >> >> >>> my >> >> >>> system using IBI. I also let program to guess all the initial >> >> >>> potential in >> >> >>> step_000. So, I do not need any external potential, but the program >> >> >>> crashes >> >> >>> in the step_001 requiring a table.xvg, I have attached my input >> >> >>> files >> >> >>> in a >> >> >>> zip file, I would be so appreciated if you could have a look and >> >> >>> help >> >> >>> me >> >> >>> overcome the problem. >> >> >> >> >> >> >> >> >> I was wrong, the table.xvg is still needed even with gromacs-2016. >> >> >> So >> >> >> just >> >> >> create a table.xvg with all zeros as y columns (or copy one over >> >> >> from >> >> >> one of >> >> >> tutorials) >> >> >> >> >> >> See the gromacs manual on details for the table format of table.xvg! >> >> >> >> >> >> Christoph >> >> >> >> >> >>> >> >> >>> >> >> >>> Thank you. >> >> >>> >> >> >>> Best regards, >> >> >>> Alex >> >> >>> >> >> >>> -- >> >> >>> You received this message because you are subscribed to the Google >> >> >>> Groups >> >> >>> "votca" group. >> >> >>> To unsubscribe from this group and stop receiving emails from it, >> >> >>> send >> >> >>> an >> >> >>> email to votca+un...@googlegroups.com. >> >> >>> To post to this group, send email to vo...@googlegroups.com. >> >> >>> Visit this group at https://groups.google.com/group/votca. >> >> >>> For more options, visit https://groups.google.com/d/optout. >> >> > >> >> > -- >> >> > You received this message because you are subscribed to the Google >> >> > Groups >> >> > "votca" group. >> >> > To unsubscribe from this group and stop receiving emails from it, >> >> > send >> >> > an >> >> > email to votca+un...@googlegroups.com. >> >> > To post to this group, send email to vo...@googlegroups.com. >> >> > Visit this group at https://groups.google.com/group/votca. >> >> > For more options, visit https://groups.google.com/d/optout. >> >> >> >> >> >> >> >> -- >> >> Christoph Junghans >> >> Web: http://www.compphys.de >> > >> > -- >> > You received this message because you are subscribed to the Google >> > Groups >> > "votca" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> > an >> > email to votca+un...@googlegroups.com. >> > To post to this group, send email to vo...@googlegroups.com. >> > Visit this group at https://groups.google.com/group/votca. >> > For more options, visit https://groups.google.com/d/optout. >> >> >> >> -- >> Christoph Junghans >> Web: http://www.compphys.de > > -- > You received this message because you are subscribed to the Google Groups > "votca" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to votca+unsubscr...@googlegroups.com. > To post to this group, send email to votca@googlegroups.com. > Visit this group at https://groups.google.com/group/votca. > For more options, visit https://groups.google.com/d/optout. -- Christoph Junghans Web: http://www.compphys.de -- You received this message because you are subscribed to the Google Groups "votca" group. To unsubscribe from this group and stop receiving emails from it, send an email to votca+unsubscr...@googlegroups.com. To post to this group, send email to votca@googlegroups.com. Visit this group at https://groups.google.com/group/votca. For more options, visit https://groups.google.com/d/optout.
