QCG is having issue with adding a large number of solvent molecules

[iconrun]

Hello,

I'm using the QCQ extension of CREST to create a water cluster (200 water molecules) around frozen stacked chromophores. The problem appears during the addition of the 93rd water molecule to the system. Below, I've attached the last lines of the output file. I was trying to change various parameters such as --wscal, -mdtime, --mddump, but the problems occurred exactly at the same step. I would like kindly ask to any help, suggestion or advice.

92 -615.535733 -16.42 -1229.06 1.362 -32.902 14.2 21.3 19243.3 normal Cluster energy not found Total Energy of cluster not found. 93 0.000000 ******* ******** 1.357 0.000 14.3 20.2 19459.7 normal

[cplett]

Hi, This might be indeed a problem of the wall potential that becomes too small. If you can send me your input coordinates, please, then I can check on this.

[iconrun]

Hi,

Thank you very much for your help. I've attached the input coordinates.

czw., 1 cze 2023 o 13:06 Christoph Plett @.***> napisał(a):

Hi, This might be indeed a problem of the wall potential that becomes too small. If you can send me your input coordinates, please, then I can check on this.

— Reply to this email directly, view it on GitHub https://github.com/crest-lab/crest/issues/202#issuecomment-1571841572, or unsubscribe https://github.com/notifications/unsubscribe-auth/BAHUK6AH5JIK2RDDTGPVYLDXJBZTBANCNFSM6AAAAAAYWU5LQA . You are receiving this because you authored the thread.Message ID: @.***>

[iconrun]

Hi,

I'm unsure whether my input coordinates were sent alongside my message. So, I've attached the coordinates below. Thank you.

71

N 4.7999300130 1.5396351138 -0.4639876729 C 3.9669080705 1.3699693428 -1.4920499080 N 4.4906072642 1.2039585851 -2.7340692178 C 2.5889661589 1.4080459635 -1.2641150651 C 2.1820955210 1.6462864919 0.0595557373 N 3.0126726256 1.8080592725 1.0911677919 C 4.2821249898 1.7254683161 0.7430452027 N 1.4877536485 1.2708198474 -2.0707838741 C 0.4639424357 1.4216589124 -1.2896636410 N 0.8139321835 1.6675822915 0.0245204573 C 0.0106741179 2.2869213090 1.0440498387 C -0.5419478998 3.6402679541 0.5564755241 C -1.8105771526 3.7858833127 1.4043804083 C -2.2844130281 2.3367880482 1.5402675520 O -1.0934650803 1.5357855551 1.4560394308 O 0.4022889073 4.6622392246 0.8037263705 O -1.4869540716 4.3027490101 2.6755120621 C -3.3053096776 1.8858653775 0.4846491971 O -2.7377793993 2.0115169809 -0.7712285114 P -3.1765560749 0.8540264640 -1.9714559630 O -4.6433663158 0.9152919654 -2.1334261780 O -2.2251948143 1.1688884565 -3.0637894373 O -2.6793576209 -0.4782050769 -1.1613995601 P -3.6317155701 -1.8124445619 -0.9838833959 O -3.9552346803 -2.4154834071 -2.2905990148 O -4.6345957300 -1.6080851590 0.0943885676 O -2.3905557393 -2.7938845387 -0.3160087579 C -2.5591959657 -3.3082405819 0.9511724666 C -1.4323141345 -2.9018220204 1.9093523161 O -0.1947202215 -3.5312545569 1.5551251543 C 0.8854732813 -2.6759128797 1.7753922675 C 0.3314373924 -1.4141933953 2.4503604005 C -1.1185196942 -1.4088657220 1.9325040827 N 1.5954280602 -2.3721635172 0.5395002693 C 2.8799885179 -1.9569787653 0.6216158973 C 3.7034679739 -1.7871875719 -0.4367504814 C 3.2315511089 -2.0667431361 -1.8468365837 C 1.8066438346 -2.5551720316 -1.8226520114 C 1.0833986974 -2.6833487713 -0.7169700712 O 0.3368086531 -1.5633348926 3.8524496778 O -1.9415596355 -0.7107168257 2.8268599258 C 5.0671537846 -1.3562402167 -0.1531707424 N 5.8933750407 -1.2369391492 -1.2411104961 O 5.5110298359 -1.1612442846 0.9704043294 H -4.2132661414 2.5054090107 0.5784714442 H -3.5820246243 0.8436640981 0.6860294804 H -2.7102534247 2.2035061736 2.5440422606 H 0.2669869600 5.3685028100 0.1648890113 H -0.6955763006 4.8533183359 2.5651430363 H -0.8106596459 3.5638888473 -0.5004007500 H -2.5647621941 4.4239258586 0.9232851214 H 0.6904662250 2.4521000648 1.8993985560 H -0.5635195934 1.3448804689 -1.6129182904 H 3.8678610778 0.9847544442 -3.4903177728 H 5.4785667174 1.0528299063 -2.8218588319 H 5.0033550321 1.8315821564 1.5432728579 H -2.5730626888 -4.4121272146 0.9093114586 H -3.5126132044 -2.9642064461 1.3807147144 H -1.8287190336 0.2162253715 2.5774752996 H -1.1613046712 -1.0106399613 0.9136852929 H -0.5273155115 -1.2197506404 4.1455258693 H 0.8964292946 -0.5155165377 2.1698107475 H 1.6030841736 -3.1680393642 2.4543087536 H -1.7019638087 -3.2261608946 2.9289947643 H 0.0884369300 -3.0997337808 -0.7299539643 H 1.3601026666 -2.8117075674 -2.7712434423 H 3.3055595585 -1.1701044817 -2.4783081226 H 3.8695439632 -2.8353607913 -2.3165402197 H 6.8038352666 -0.8361831807 -1.0868364793 H 5.5436678582 -1.2861611472 -2.1790591573 H 3.2789007067 -1.8013067816 1.6166767756

[cplett]

Hi, I tested the system with 200 water molecules and it seemed to work for me: I guess that this might be an issue of an older version. Can you, please, tell me if you use a crest version you compiled from the source code or a release and if you use the old xtbiff or the new aISS algorithm of xtb to grow the cluster? Then, I can check on the specific version.

[iconrun]

Hi,

Thank you for sending the results.

According to the visualisation of the system, I guess that the NADH molecule was not constrained during the adding water molecules. The final NADH structure surrounded by 200 water molecules is quite far from the starting structure, which I must maintain during the cluster growth. To constrain the NADH molecule, I added '--cinp constrain' to the command line. Could it be a problem for growing the cluster?

The contents of the constrain file: $constrain atoms: 1-71 $end

I use the release versions of a crest (2.12), xtb (6.6.0), and xtbiff (1.1) downloaded from the github. According to my understanding, I've been using only the crest with the xtbiff algorithm. Is there a way to force using the new aISS algorithm of xtb to grow the cluster?

Thank you in advance for your help.

[cplett]

Hi, Currently, only a crest version build with the source code is able to use the aISS. This will be the default then. Regarding the fixed solute, you can use the --fixsolute and --nopreopt flags to keep the solute fixed during the growth. For the ensemble you will indeed require the xcontrol. I will check this and also the version with the xtbiff.

[pprcht]

@cplett actually the continous release should include the aISS usage capability, shouldn't it?

[cplett]

Somehow, the continuous release didn't update after the last commits.

[pprcht]

According to the description the release has the status of commit #189 of April 3rd, so it should be up to date. You can clearly see the binaries were uploaded on that date. The release tag never changes from Dec 23, 2022.

[iconrun]

Thank you for all your suggestions.

I tested the aISS algorithm for the cluster growth from the continuous release, and it has the same problem as the xtbiff. When I freeze the solute (using the .xcontrol file and adding the --fixsolute and --nopreopt flags), I cannot add more than 94 water molecules to the system. Formally, the NADH molecule has a charge of -2. Manipulating the charge, I wanted to check whether the negatively charged system could affect the cluster growth, but it's not.

Any advice would be appreciated.

[cplett]

Hi, I tested a little bit more (also with fixed solute and xcontrol file), but I didn't observe a crash of the program. For further evaluation, it would be great if you could provide me the whole printout of crest and maybe also the tmp_grow directory in the qcg_tmp, please.

[iconrun]

Hi,

I've attached the whole printout of crest. Thank you. [nadh-crest-qcg.zip](https://github.com/crest-lab/crest/files/11743133/nadh-crest-qcg.zip

[cplett]

Hi, It seems that a xtb optimization somehow crashed. Did you give the program enough stack size by configuring the .bashrc (if you are working on a Linux computer? Helpful settings are: ulimit -s unlimited export OMP_STACKSIZE=4G export OMP_NUM_THREADS=,1 Further instructions can be found here.

[iconrun]

Hi,

Thank you for the suggestions. Increasing the value of the OMP_STACKSIZE variable helped. I tested both the aISS and xtbiff approaches, and everything works.

[github-actions[bot]]

This issue had no activity for 6 months. It will be closed in 1 week unless there is some new activity.