Dear Christoph, Thank you so much for the additional tips. Sincerely, Zidan
On Tuesday, May 10, 2016 at 1:45:56 AM UTC+2, Christoph Junghans wrote: > > > > 2016-05-09 15:13 GMT-06:00 Zidan Zhang <[email protected] <javascript:>>: > >> Dear Christoph, >> Thank you very much for your reply! >> g_rdf and csg_stat indeed give the identical results. >> As shown in the attached figure, I have defined several dihedrals, but >> even I have 40000 frames (totally run 200 ns), the distributions don't look >> well, so I decided not to use dihedrals. But the RDFs were generated with >> the setting.xml file in which the dihedrals were defined. When doing IBI >> for nonbonded interactions, I wrote a <map> tag in the setting.xml file in >> which I have deleted these dihedrals (as shown in the earlier attachments), >> so that's why I have different results and the nonbonded interactions don't >> converge. Now I know the reason and IBI works fine. It's my mistake. >> Thank you again! >> > That is good news! > > Beside doing bonded IBI there are only two cases when you want to use a > mapping file: 1.) adding extra exclusion, which can also be done by using a > special tpr file for the rdf calculation (in xml: > cg.inverse.gromacs.rdf.topol). 2.) Make csg_stat calculate bonded > distributions so that you can check with get perturbed. > > Christoph > >> Best, >> Zidan >> >> >> <https://lh3.googleusercontent.com/-NYso0_SMnl4/VzD7iFug1UI/AAAAAAAAANQ/9k9r2ToB7Jc8pi8LfSVsmLfH3UEIYly9ACLcB/s1600/dihedral.png> >> >> >> On Sunday, May 8, 2016 at 10:44:52 PM UTC+2, Christoph Junghans wrote: >>> >>> >>> >>> 2016-05-08 12:07 GMT-06:00 Zidan Zhang <[email protected]>: >>> >>>> Dear all, >>>> I think I find the problem, it's the different between g_rdf and the >>>> csg_stat, after reading several threads, I think it will converge. When I >>>> get the new data, I will update this thread. Thank you! >>>> >>> g_rdf and csg_stat should give identical results if they don't than it >>> is usually due to missing exclusions. >>> However, if you are just doing IBI on non-bonded interactions (and not >>> the bonded interactions), you don't need a 1:1 mapping file and VOTCA will >>> read the exclusion directly from the tpr file. >>> >>> Christoph >>> >>> Best, >>>> Zidan >>>> >>>> >>>> On Sunday, May 8, 2016 at 7:12:42 PM UTC+2, Zidan Zhang wrote: >>>>> >>>>> Dear Denis, >>>>> Thank you for the reply, I have deleted the xtc file as listed in >>>>> <cleanlist> section, I will re-run the last iteration to check RMSD. >>>>> Sincerely, >>>>> Zidan >>>>> >>>>> On Sunday, May 8, 2016 at 7:03:32 PM UTC+2, Denis Andrienko wrote: >>>>>> >>>>>> Hi, >>>>>> >>>>>> IBI requires canonical sampling of the coarse-grained radial >>>>>> distribution function. >>>>>> Can you check if your system is still in a liquid state (i.e. not in >>>>>> a glassy one)? You can do it by calculating RMSD of last iterations and >>>>>> checking if a molecule diffused more than its own size. If not, you >>>>>> might >>>>>> want to increase the runtime of every IBI step. >>>>>> >>>>>> In some exotic cases you just have to do more IBI steps. >>>>>> >>>>>> Best, >>>>>> Denis >>>>>> >>>>>> On Sunday, May 8, 2016 at 3:53:32 PM UTC+2, Zidan Zhang wrote: >>>>>>> >>>>>>> Dear all, >>>>>>> When optimise the nonbonded interactions, some of them don't >>>>>>> converge, like shown in the picture. The system is a bulk system and >>>>>>> has no >>>>>>> crystal. >>>>>>> I have tried several solutions: >>>>>>> 1. change the scale factor (1.0, 0.7, 0.5); >>>>>>> 2. change the initial guess by using high temperature (default 458 >>>>>>> K, high T=600 K) in the setting.xml (nonbonded.xml as attached); >>>>>>> 3. the order of optimisation. >>>>>>> All of them have the same results. >>>>>>> How can I solve this problem? >>>>>>> Best regards, >>>>>>> Zidan >>>>>>> >>>>>>> >>>>>>> >>>>>>> <https://lh3.googleusercontent.com/-eOfyHp6X1bg/Vy9Cx2MzKVI/AAAAAAAAANA/wuLn1LkrZOki0EIe1Gj2KyekymfeQmrjACLcB/s1600/rb-rb.png> >>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>> 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 [email protected]. >>>> To post to this group, send email to [email protected]. >>>> 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 [email protected] <javascript:>. >> To post to this group, send email to [email protected] <javascript:> >> . >> 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 [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/votca. For more options, visit https://groups.google.com/d/optout.
