On Thu, May 16, 2013 at 2:04 PM, bharat gupta <bharat.85.m...@gmail.com>wrote:
> Dear Sir, > > Here's the result of three different runs : > > Temperature distribution for three trials > > Repeat-1 280 298 317 337 359 382 406 432 460 489 520 554 589 627 > Repeat-2 280 299 319 340 363 388 414 441 471 503 536 572 611 > Repeat-3 280 300 322 345 370 397 426 457 490 526 564 605 649 > > md.log files output from three different trials: > > Repeat-1 .37 .28 .26 .30 .25 .29 .32 .35 .32 .35 .36 .32 .31 > Repeat-2 .30 .33 .30 .25 .19 .27 .30 .31 .27 .40 .34 .31 > Repeat-3 .18 .22 .26 .34 .26 .28 .25 .27 .27 .25 .27 .22 > > I think as the required acceptance value all the three trials are fine, but > trail 3 would be much better to continue the further runs and anlysis ?? > Probably. But exchange acceptance is a poor proxy for sampling efficiency - see recent discussions of REMD on this list. Mark So, is it fine to continue with the third simulation ?? But still the > problem is that I am not getting the exact graphs with xmgrace?? > > > On Thu, May 16, 2013 at 5:36 PM, XAvier Periole <x.peri...@rug.nl> wrote: > > > > > You have to convince yourself, not me :)) But I can give you my opinion … > > > > On May 16, 2013, at 10:33 AM, bharat gupta <bharat.85.m...@gmail.com> > > wrote: > > > > > Okay Sir, I will try two-three combinations this time and will report > > back > > > to you ... > > > > > > > > > On Thu, May 16, 2013 at 5:25 PM, XAvier Periole <x.peri...@rug.nl> > > wrote: > > > > > >> > > >> An acceptance ratio of 0.2/0.3 is normally best. The problem with high > > >> acceptance ratio is that it means that a large portion of the > exchanges > > are > > >> just back and forth exchanges between consecutive exchange and are > thus > > >> disturbing the system more that actually helping sampling. > > >> > > >> I do not know particularly the paper you mention but if you like what > > they > > >> do, it is your choice at the end. > > >> > > >> Why don;t you just increase the spacing between the replicas? You will > > >> need less replicas and potentially you could run two simulations > > instead of > > >> one and evaluate the convergence ... > > >> > > >> On May 16, 2013, at 1:50 AM, bharat gupta <bharat.85.m...@gmail.com> > > >> wrote: > > >> > > >>> The plots that I showed in my last mail were for all replicas. I > tried > > >>> plotting the first 500 ps of replica_index and replica_time files. I > > >> think > > >>> the plots look fine, and there could be problem with the plotting > tool > > . > > >>> Here the link for both files , > > >>> https://www.dropbox.com/s/2g16mlxfsme4rx2/replica_temp.bmp > > >>> https://www.dropbox.com/s/8jfs0b9whu6j7lo/replica_index.bmp > > >>> > > >>> Now regarding the high acceptance ratio which is 0.5 , I came across > a > > >>> paper (http://www.pnas.org/content/100/13/7587.full.pdf), here they > > have > > >>> mentioned that their average acceptance ratio ranged between 30 to > > 80%. I > > >>> have a question here, how did they calculate the range for the > average > > >>> acceptance ratio or is it average ratio for each replica . Actually, > > this > > >>> is the reference I am following. I am also interested in peptide > > folding > > >>> simulation, similar to this article. > > >>> > > >>> I want to know, whether the average acceptance ratio that I have got > > for > > >> my > > >>> trial simulation is correct , together with the replica_temp and > > >>> replica_remd plots. Can I proceed for large production runs to > complete > > >> my > > >>> experiment ?? > > >>> > > >>> > > >>> > > >>> > > >>> On Tue, May 14, 2013 at 6:34 PM, XAvier Periole <x.peri...@rug.nl> > > >> wrote: > > >>> > > >>>> > > >>>> The interval between the exchange trial affect the efficiency of > REMD > > >> but > > >>>> not the the exchange ratio (at least in principle). > > >>>> > > >>>> In you case I am not sure what the plot are showing! Are these > showing > > >> all > > >>>> the replicas? what are the units? > > >>>> > > >>>> On May 14, 2013, at 5:07 AM, bharat gupta <bharat.85.m...@gmail.com > > > > >>>> wrote: > > >>>> > > >>>>> Dear Sir, > > >>>>> > > >>>>> Here's the result for the REMD trial with large temperature gaps. > > >>>>> > > >>>>> Temp. distribution : 280.0 294.9 310.7 327.3 344.7 363.1 382.5 > 402.9 > > >>>> 424.4 > > >>>>> 447.1 471.0 496.1 522.6 550.5 579.9 610.8 > > >>>>> > > >>>>> Out of md16.log : > > >>>>> > > >>>>> Replica exchange statistics > > >>>>> Repl 249 attempts, 125 odd, 124 even > > >>>>> Repl average probabilities: > > >>>>> Repl 0 1 2 3 4 5 6 7 8 9 10 11 > > 12 > > >>>>> 13 14 15 > > >>>>> Repl .40 .34 .38 .43 .43 .36 .45 .40 .37 .48 .47 > .45 > > >> .47 > > >>>>> .44 .46 > > >>>>> > > >>>>> Repl number of exchanges: > > >>>>> Repl 0 1 2 3 4 5 6 7 8 9 10 11 > > 12 > > >>>>> 13 14 15 > > >>>>> Repl 50 42 46 52 57 40 58 49 42 53 61 > 63 > > >> 56 > > >>>>> 57 58 > > >>>>> > > >>>>> Repl average number of exchanges: > > >>>>> Repl 0 1 2 3 4 5 6 7 8 9 10 11 > > 12 > > >>>>> 13 14 15 > > >>>>> Repl .40 .34 .37 .42 .46 .32 .46 .40 .34 .43 .49 > .51 > > >> .45 > > >>>>> .46 .46 > > >>>>> Average acceptance ratio : 0.46 > > >>>>> > > >>>>> But, the repli_index.xvg and replica_temp.xvg files still shows > that > > >> the > > >>>>> replicas does not exchange equally well . > > >>>>> > > >>>>> https://www.dropbox.com/s/zkbwpuj7l2o282b/replica_index.png > > >>>>> https://www.dropbox.com/s/0c8gp584v1hvlbx/replica_temp.png > > >>>>> > > >>>>> what could be wrong in this case?? Is it the mdp file settings or > > >>>> implicit > > >>>>> solvent setting. Does the time to replica to exhange also affects > > their > > >>>>> swapping ?? > > >>>>> > > >>>>> > > >>>>> > > >>>>> On Tue, May 14, 2013 at 12:24 AM, XAvier Periole <x.peri...@rug.nl > > > > >>>> wrote: > > >>>>> > > >>>>>> > > >>>>>> You need to increase the temperature gaps indeed if you want > > >> acceptance > > >>>>>> ratio ~0.2/0.3. But again this won't work with the water … > > >>>>>> > > >>>>>> It is not clear what happens in your index file but probably a > > problem > > >>>>>> from grace to plot so many points … you can try to increase the > "Max > > >>>>>> drawing path length" in the preference menu of grace. > > >>>>>> > > >>>>>> On May 13, 2013, at 4:22 PM, bharat gupta < > bharat.85.m...@gmail.com > > > > > >>>>>> wrote: > > >>>>>> > > >>>>>>> Dear Sir, > > >>>>>>> > > >>>>>>> I repeated the simulation again for 25 replicas with the > following > > >>>> temp. > > >>>>>>> distribution . > > >>>>>>> > > >>>>>>> 280 > > >>>>>>> 289.1 > > >>>>>>> 298.5 > > >>>>>>> 308.2 > > >>>>>>> 318.2 > > >>>>>>> 328.6 > > >>>>>>> 339.3 > > >>>>>>> 350.3 > > >>>>>>> 361.7 > > >>>>>>> 373.5 > > >>>>>>> 385.6 > > >>>>>>> 398.1 > > >>>>>>> 411.1 > > >>>>>>> 424.4 > > >>>>>>> 438.3 > > >>>>>>> 452.5 > > >>>>>>> 467.2 > > >>>>>>> 482.4 > > >>>>>>> 498.1 > > >>>>>>> 514.3 > > >>>>>>> 531.0 > > >>>>>>> 548.3 > > >>>>>>> 566.1 > > >>>>>>> 584.5 > > >>>>>>> 603.5 > > >>>>>>> 623.2 > > >>>>>>> > > >>>>>>> The output of md.log file is :- > > >>>>>>> > > >>>>>>> Replica exchange statistics > > >>>>>>> Repl 24999 attempts, 12500 odd, 12499 even > > >>>>>>> Repl average probabilities: > > >>>>>>> Repl 0 1 2 3 4 5 6 7 8 9 10 11 > > >> 12 > > >>>>>>> 13 14 15 16 17 18 19 20 21 22 23 24 25 > > >>>>>>> Repl .63 .63 .62 .62 .61 .61 .60 .60 .59 .59 .58 > > .59 > > >>>> .59 > > >>>>>>> .60 .60 .61 .62 .62 .63 .64 .64 .65 .65 .66 .66 > > >>>>>>> > > >>>>>>> Repl number of exchanges: > > >>>>>>> Repl 0 1 2 3 4 5 6 7 8 9 10 11 > > >> 12 > > >>>>>>> 13 14 15 16 17 18 19 20 21 22 23 24 25 > > >>>>>>> Repl 7822 7752 7816 7760 7639 7628 7511 7442 7375 7332 7312 > > 7424 > > >>>> 7408 > > >>>>>>> 7410 7522 7559 7684 7697 7878 7927 7917 8073 8151 8208 8266 > > >>>>>>> > > >>>>>>> Repl average number of exchanges: > > >>>>>>> Repl 0 1 2 3 4 5 6 7 8 9 10 11 > > >> 12 > > >>>>>>> 13 14 15 16 17 18 19 20 21 22 23 24 25 > > >>>>>>> Repl .63 .62 .63 .62 .61 .61 .60 .60 .59 .59 .58 > > .59 > > >>>> .59 > > >>>>>>> .59 .60 .60 .61 .62 .63 .63 .63 .65 .65 .66 .66 > > >>>>>>> > > >>>>>>> The average acceptance ration is around 0.6 which is still high. > > >>>>>>> > > >>>>>>> The link for replica_temp,replica_index : > > >>>>>>> https://www.dropbox.com/s/c7soajnwc3uww8j/replica_temp.png > > >>>>>>> https://www.dropbox.com/s/wvx82m4c6cnsfit/replica_index.png > > >>>>>>> > > >>>>>>> The temp files look better but the index file looks weird ... > > >>>>>>> > > >>>>>>> Do i need to experiment with the gap difference in order to get > the > > >>>>>>> required ration of 0.2-0.3 ?? There is some problem with the .mdp > > >> file > > >>>>>>> settings?? > > >>>>>>> > > >>>>>>> -- > > >>>>>>> Bharat > > >>>>>>> -- > > >>>>>>> gmx-users mailing list gmx-users@gromacs.org > > >>>>>>> http://lists.gromacs.org/mailman/listinfo/gmx-users > > >>>>>>> * Please search the archive at > > >>>>>> http://www.gromacs.org/Support/Mailing_Lists/Search before > posting! > > >>>>>>> * Please don't post (un)subscribe requests to the list. Use the > > >>>>>>> www interface or send it to gmx-users-requ...@gromacs.org. > > >>>>>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > >>>>>> > > >>>>>> -- > > >>>>>> gmx-users mailing list gmx-users@gromacs.org > > >>>>>> http://lists.gromacs.org/mailman/listinfo/gmx-users > > >>>>>> * Please search the archive at > > >>>>>> http://www.gromacs.org/Support/Mailing_Lists/Search before > posting! > > >>>>>> * Please don't post (un)subscribe requests to the list. Use the > > >>>>>> www interface or send it to gmx-users-requ...@gromacs.org. > > >>>>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > >>>>>> > > >>>>> -- > > >>>>> gmx-users mailing list gmx-users@gromacs.org > > >>>>> http://lists.gromacs.org/mailman/listinfo/gmx-users > > >>>>> * Please search the archive at > > >>>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting! > > >>>>> * Please don't post (un)subscribe requests to the list. Use the > > >>>>> www interface or send it to gmx-users-requ...@gromacs.org. > > >>>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > >>>> > > >>>> -- > > >>>> gmx-users mailing list gmx-users@gromacs.org > > >>>> http://lists.gromacs.org/mailman/listinfo/gmx-users > > >>>> * Please search the archive at > > >>>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting! > > >>>> * Please don't post (un)subscribe requests to the list. Use the > > >>>> www interface or send it to gmx-users-requ...@gromacs.org. > > >>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > >>>> > > >>> -- > > >>> gmx-users mailing list gmx-users@gromacs.org > > >>> http://lists.gromacs.org/mailman/listinfo/gmx-users > > >>> * Please search the archive at > > >> http://www.gromacs.org/Support/Mailing_Lists/Search before posting! > > >>> * Please don't post (un)subscribe requests to the list. Use the > > >>> www interface or send it to gmx-users-requ...@gromacs.org. > > >>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > >> > > >> -- > > >> gmx-users mailing list gmx-users@gromacs.org > > >> http://lists.gromacs.org/mailman/listinfo/gmx-users > > >> * Please search the archive at > > >> http://www.gromacs.org/Support/Mailing_Lists/Search before posting! > > >> * Please don't post (un)subscribe requests to the list. Use the > > >> www interface or send it to gmx-users-requ...@gromacs.org. > > >> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > >> > > > -- > > > gmx-users mailing list gmx-users@gromacs.org > > > http://lists.gromacs.org/mailman/listinfo/gmx-users > > > * Please search the archive at > > http://www.gromacs.org/Support/Mailing_Lists/Search before posting! > > > * Please don't post (un)subscribe requests to the list. Use the > > > www interface or send it to gmx-users-requ...@gromacs.org. > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > > > -- > > gmx-users mailing list gmx-users@gromacs.org > > http://lists.gromacs.org/mailman/listinfo/gmx-users > > * Please search the archive at > > http://www.gromacs.org/Support/Mailing_Lists/Search before posting! > > * Please don't post (un)subscribe requests to the list. Use the > > www interface or send it to gmx-users-requ...@gromacs.org. > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > > -- > gmx-users mailing list gmx-users@gromacs.org > http://lists.gromacs.org/mailman/listinfo/gmx-users > * Please search the archive at > http://www.gromacs.org/Support/Mailing_Lists/Search before posting! > * Please don't post (un)subscribe requests to the list. Use the > www interface or send it to gmx-users-requ...@gromacs.org. > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > -- gmx-users mailing list gmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
