Re: Fwd: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer
By the way I also wounder to know how I could keep my system after equilibration. E.g I've equilibrated CCl4+water system and obtain stabile system on desired timescale. Than I've used genbox to merge my pre-oriented peptide with the pre-equilibrated system. genbox -cp peptide_newbox.gro -cs CCl4_water_box.gro -o peptide_in_box.gro As the consequence I've obtained system with peptide inserted in CCl4 but without water layer. Is there any way to preserve water layer during Genbox peptide insertion? James -- gmx-users mailing listgmx-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
Re: Fwd: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer
Mark, I'm using exact all parameters wich I found in different experimental work. By the way reducing of integration step to 1fs provide me with better equilibration of the Ccl4 system ( I've being obtained stabile system during 3 ns) but I had a problems during ntp equilibration when I inserted test peptide into that pre-equilibrated Ccl4 system and made two surrounded layer of water. My system was quickly eqxanded on Z-dimension and slightly shrinked on X. I think that such problem could be due to some problems with the vdw radius value for CCl4. E.g I didnt find this value in the vdwradii.dat file. James 2012/2/16 Mark Abraham mark.abra...@anu.edu.au On 16/02/2012 1:45 AM, James Starlight wrote: Mark, I've used that dimensions in accordance to some literature where the same membrane-mimicking simulation were performed. I've tried to rise cutoffs Don't, that breaks your model physics and makes it even more likely you will encounter problems with the system dimensions becoming too small for the cut-off! and dicrease integration step but my system have been stil crashed during npt. I'm using pcoupl= Parrinello-Rahman wich I've found in the KALP tutorial because I have not found the same npt example file in the Biphastic tutorial :) So you're following some other work and not copying their equilibration protocol and/or model physics? Could you advise me another p_coup algorithm for my Ccl4 system? There's only two choices available. Manual 3.4.9 specifically warns against one of them for equilibration. What is there to say? You should be sure to construct a simple case and get the model physics validated. For the moment, forget about all the stuff where you were struggling to insert more CCl4 into a box with CCl4 (probably creating a far-from-equilibrium starting configuration). Don't try to learn to run on stilts while shaving. Learn to shave, then to walk on stilts, then to run, then start combining them. Mark James -- Forwarded message -- From: Mark Abraham mark.abra...@anu.edu.au Date: 2012/2/15 Subject: Re: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer To: Discussion list for GROMACS users gmx-users@gromacs.org On 15/02/2012 4:45 PM, James Starlight wrote: Mark, due to hight density the volume of my system have been slightly increased and during NPT phase I've obtained error Fatal error: One of the box vectors has become shorter than twice the cut-off length or box_yy-|box_zy| or box_zz has become smaller than the cut-off. I'm using 0.9 for electrostatic and 1.4 for vdw cutofs and the dimensions of my box was 6.5 3 3 on the initial step and 6.6 3 3.3 before crush :) I want prevent such expansion of my system by increasing of pressure and/ or compressibility but I have not found exact sollution yet. Your system is dangerously small for those cut-offs if your initial density is not correct for your model physics. Your y and z dimensions only just contain a full cut-off sphere. You should also make sure you are following the advice about choice of P-coupling algorithm in manual 3.4.9, and consider using a very small integration time step. I remain unconvinced by this thread that you have generated a starting configuration that does not have atomic clashes. Mark James 2012/2/14 Mark Abraham mark.abra...@anu.edu.au On 14/02/2012 11:01 PM, James Starlight wrote: This also was solved by the some extra minimisation steps. I've forced with another problem :D During npt equilibration my system have slightly expanded so my desired volume and density were perturbed. I've noticed the below options in npt wich could help me ref_p= 1 1 compressibility = 4.5e-5 i'm using this compressibility value because I'm modelling the lipid-like environment so I think that I must increase pressure. Could you remind me the dependence of pressure from density and volume for liquids ? :) Your forcefield, simulation cell contents and .mdp settings will determine the equilibrium density. Whether you need to do anything depends on whether you've made a statistically significant post-equilibration measurement of your average density. Haphazardly increasing the reference pressure for the coupling will reduce the volume, but now you are simulating at that pressure. See http://www.gromacs.org/Documentation/Terminology/Pressure for background info. Mark James 2012/2/14 James Starlight jmsstarli...@gmail.com It seems that I've fixed that problem by reduce vdv radii for Cl during defining of my box Eventually I've obtained box with the desired density than I've delete vdvradii.dat for my wor dir by when I've launched equilibration I've oibtained Fatal error: Too many LINCS warnings (1598) If you know what you are doing you can adjust the lincs warning threshold in your mdp file I've never seen this before I'm using 1.o
Re: Fwd: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer
So my problem still is 1) During npt without peptide my Ccl4 system expan rapidly in X-dimension. That produce error on 2nd ns of equilibration One of the box vectors has become shorter than twice the cut-off length or box_yy-|box_zy| or box_zz has become smaller than the cut-off. 2) During npt with peptide my system expan rapidly in Z direction ( along that axe my peptide is oriented) So as concequence I've obtained the same error. Possible sollution wich I've found in litterature 1) To fix one dimension of my PB. How I could fix the X dimension of my box during npt phace of pure Ccl4 and allow only change in 2 other dimensions ? 2) I have not found radius for Cl for Gromos56 ff. Is this possible to decrease that value during npt phase to obtain system with desired hight density? What value should I use? James 2012/2/16 James Starlight jmsstarli...@gmail.com Mark, I'm using exact all parameters wich I found in different experimental work. By the way reducing of integration step to 1fs provide me with better equilibration of the Ccl4 system ( I've being obtained stabile system during 3 ns) but I had a problems during ntp equilibration when I inserted test peptide into that pre-equilibrated Ccl4 system and made two surrounded layer of water. My system was quickly eqxanded on Z-dimension and slightly shrinked on X. I think that such problem could be due to some problems with the vdw radius value for CCl4. E.g I didnt find this value in the vdwradii.dat file. James 2012/2/16 Mark Abraham mark.abra...@anu.edu.au On 16/02/2012 1:45 AM, James Starlight wrote: Mark, I've used that dimensions in accordance to some literature where the same membrane-mimicking simulation were performed. I've tried to rise cutoffs Don't, that breaks your model physics and makes it even more likely you will encounter problems with the system dimensions becoming too small for the cut-off! and dicrease integration step but my system have been stil crashed during npt. I'm using pcoupl= Parrinello-Rahman wich I've found in the KALP tutorial because I have not found the same npt example file in the Biphastic tutorial :) So you're following some other work and not copying their equilibration protocol and/or model physics? Could you advise me another p_coup algorithm for my Ccl4 system? There's only two choices available. Manual 3.4.9 specifically warns against one of them for equilibration. What is there to say? You should be sure to construct a simple case and get the model physics validated. For the moment, forget about all the stuff where you were struggling to insert more CCl4 into a box with CCl4 (probably creating a far-from-equilibrium starting configuration). Don't try to learn to run on stilts while shaving. Learn to shave, then to walk on stilts, then to run, then start combining them. Mark James -- Forwarded message -- From: Mark Abraham mark.abra...@anu.edu.au Date: 2012/2/15 Subject: Re: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer To: Discussion list for GROMACS users gmx-users@gromacs.org On 15/02/2012 4:45 PM, James Starlight wrote: Mark, due to hight density the volume of my system have been slightly increased and during NPT phase I've obtained error Fatal error: One of the box vectors has become shorter than twice the cut-off length or box_yy-|box_zy| or box_zz has become smaller than the cut-off. I'm using 0.9 for electrostatic and 1.4 for vdw cutofs and the dimensions of my box was 6.5 3 3 on the initial step and 6.6 3 3.3 before crush :) I want prevent such expansion of my system by increasing of pressure and/ or compressibility but I have not found exact sollution yet. Your system is dangerously small for those cut-offs if your initial density is not correct for your model physics. Your y and z dimensions only just contain a full cut-off sphere. You should also make sure you are following the advice about choice of P-coupling algorithm in manual 3.4.9, and consider using a very small integration time step. I remain unconvinced by this thread that you have generated a starting configuration that does not have atomic clashes. Mark James 2012/2/14 Mark Abraham mark.abra...@anu.edu.au On 14/02/2012 11:01 PM, James Starlight wrote: This also was solved by the some extra minimisation steps. I've forced with another problem :D During npt equilibration my system have slightly expanded so my desired volume and density were perturbed. I've noticed the below options in npt wich could help me ref_p= 1 1 compressibility = 4.5e-5 i'm using this compressibility value because I'm modelling the lipid-like environment so I think that I must increase pressure. Could you remind me the dependence of pressure from density and volume for liquids ? :) Your forcefield, simulation cell contents and .mdp settings will determine the
Re: Fwd: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer
On 17/02/2012 4:46 PM, James Starlight wrote: So my problem still is 1) During npt without peptide my Ccl4 system expan rapidly in X-dimension. That produce error on 2nd ns of equilibration One of the box vectors has become shorter than twice the cut-off length or box_yy-|box_zy| or box_zz has become smaller than the cut-off. 2) During npt with peptide my system expan rapidly in Z direction ( along that axe my peptide is oriented) So as concequence I've obtained the same error. What does a cube of CCl4 do? If it is not stable at an acceptable density then your parameters don't work at all. Starting with simple things and adding complexity gradually is always a good policy, even when you're an expert. There's so many things that can be slightly wrong... Possible sollution wich I've found in litterature 1) To fix one dimension of my PB. How I could fix the X dimension of my box during npt phace of pure Ccl4 and allow only change in 2 other dimensions ? See pcoupltype in manual 7.3.15 2) I have not found radius for Cl for Gromos56 ff. Is this possible to decrease that value during npt phase to obtain system with desired hight density? What value should I use? As someone else said, the radius is only material during things like genbox. In the simulation, your nonbonded parameters are the important thing. Mark -- gmx-users mailing listgmx-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
Re: Fwd: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer
Mark, The pure Ccl4 cube was expanded in X dimension during npt phase. By the way I've been able to prevent it by increasing the ref_p of X up to 5 ussing semiisotropic pcoupltype By when I've inserted peptide on the npt of that system my Ccl4 box was expanded on Z rapidly. Might this way to increase ref_p be usefull ? I want to obtain very tightly packet Ccl4 layer with high density (~1500kg/m^3) to mimick membrane acyl-chain hydrophobic layer James -- gmx-users mailing listgmx-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
Re: Fwd: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer
On 17/02/2012 5:33 PM, James Starlight wrote: Mark, The pure Ccl4 cube was expanded in X dimension during npt phase. So assuming you simulated long enough to have reasonable convergence, your model is not stable at the initial volume with whatever that reference pressure was. By the way I've been able to prevent it by increasing the ref_p of X up to 5 ussing semiisotropic pcoupltype By when I've inserted peptide on the npt of that system my Ccl4 box was expanded on Z rapidly. It's likely a waste of time to insert stuff into a solvent system that is far from equilibrium. Might this way to increase ref_p be usefull ? Depends what you want to measure under what conditions. I want to obtain very tightly packet Ccl4 layer with high density (~1500kg/m^3) to mimick membrane acyl-chain hydrophobic layer So either you'll need high pressure or a different model of CCl4. Surely the work you are following addressed this. Your decision. Mark -- gmx-users mailing listgmx-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
Re: Fwd: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer
On 16/02/2012 1:45 AM, James Starlight wrote: Mark, I've used that dimensions in accordance to some literature where the same membrane-mimicking simulation were performed. I've tried to rise cutoffs Don't, that breaks your model physics and makes it even more likely you will encounter problems with the system dimensions becoming too small for the cut-off! and dicrease integration step but my system have been stil crashed during npt. I'm using pcoupl= Parrinello-Rahman wich I've found in the KALP tutorial because I have not found the same npt example file in the Biphastic tutorial :) So you're following some other work and not copying their equilibration protocol and/or model physics? Could you advise me another p_coup algorithm for my Ccl4 system? There's only two choices available. Manual 3.4.9 specifically warns against one of them for equilibration. What is there to say? You should be sure to construct a simple case and get the model physics validated. For the moment, forget about all the stuff where you were struggling to insert more CCl4 into a box with CCl4 (probably creating a far-from-equilibrium starting configuration). Don't try to learn to run on stilts while shaving. Learn to shave, then to walk on stilts, then to run, then start combining them. Mark James -- Forwarded message -- From: *Mark Abraham* mark.abra...@anu.edu.au mailto:mark.abra...@anu.edu.au Date: 2012/2/15 Subject: Re: [gmx-users] Npt equilibration of the membrane-mimicking CCl4 layer To: Discussion list for GROMACS users gmx-users@gromacs.org mailto:gmx-users@gromacs.org On 15/02/2012 4:45 PM, James Starlight wrote: Mark, due to hight density the volume of my system have been slightly increased and during NPT phase I've obtained error Fatal error: One of the box vectors has become shorter than twice the cut-off length or box_yy-|box_zy| or box_zz has become smaller than the cut-off. I'm using 0.9 for electrostatic and 1.4 for vdw cutofs and the dimensions of my box was 6.5 3 3 on the initial step and 6.6 3 3.3 before crush :) I want prevent such expansion of my system by increasing of pressure and/ or compressibility but I have not found exact sollution yet. Your system is dangerously small for those cut-offs if your initial density is not correct for your model physics. Your y and z dimensions only just contain a full cut-off sphere. You should also make sure you are following the advice about choice of P-coupling algorithm in manual 3.4.9, and consider using a very small integration time step. I remain unconvinced by this thread that you have generated a starting configuration that does not have atomic clashes. Mark James 2012/2/14 Mark Abraham mark.abra...@anu.edu.au mailto:mark.abra...@anu.edu.au On 14/02/2012 11:01 PM, James Starlight wrote: This also was solved by the some extra minimisation steps. I've forced with another problem :D During npt equilibration my system have slightly expanded so my desired volume and density were perturbed. I've noticed the below options in npt wich could help me ref_p= 1 1 compressibility = 4.5e-5 i'm using this compressibility value because I'm modelling the lipid-like environment so I think that I must increase pressure. Could you remind me the dependence of pressure from density and volume for liquids ? :) Your forcefield, simulation cell contents and .mdp settings will determine the equilibrium density. Whether you need to do anything depends on whether you've made a statistically significant post-equilibration measurement of your average density. Haphazardly increasing the reference pressure for the coupling will reduce the volume, but now you are simulating at that pressure. See http://www.gromacs.org/Documentation/Terminology/Pressure for background info. Mark James 2012/2/14 James Starlight jmsstarli...@gmail.com mailto:jmsstarli...@gmail.com It seems that I've fixed that problem by reduce vdv radii for Cl during defining of my box Eventually I've obtained box with the desired density than I've delete vdvradii.dat for my wor dir by when I've launched equilibration I've oibtained Fatal error: Too many LINCS warnings (1598) If you know what you are doing you can adjust the lincs warning threshold in your mdp file I've never seen this before I'm using 1.o cutoff for pme and 1.4 for vdv my LINKS parameters are ; Bond parameters continuation= no; first dynamics run constraint_algorithm = lincs; holonomic constraints constraints= all-bonds; all bonds (even heavy atom-H bonds) constrained lincs_iter= 1; accuracy of LINCS lincs_order= 4; also related to
