Indeed the Repeat-3 seems good. But I would guess you did not run too long, right! That would explain the distribution of values!
On 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 ?? > > 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
