normal Incorrect density

  • edroaldo
  • edroaldo's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 5 months ago - 7 years 5 months ago #859 by edroaldo
Incorrect density was created by edroaldo
Dear,

I did a simulation in a system with a lipid bilayer (1200 lipids), ions, water and one nanoparticle. The box dimensions are 20x20x18 and I add about 49000 CG water beads in this system. After a 50ns simulation, I use the gromacs tool g_density to calc the density of some molecules in my system. The density for water was above of 1000 kg/m³ and the papers that I am following show results below to this value. One of this papers says that about 60000 CG water beads was added in a system that contains 1152 lipids. Someone please give me some direction? Do I need remove water from my system?

Thank you very much.
Best regards,
Edroaldo.

Please Log in or Create an account to join the conversation.

More
7 years 5 months ago #862 by djurre
Replied by djurre on topic Incorrect density
Dear Edroaldo,
I'm guessing your taking the maximum peak hight from the density profile to be the water density. Are you averaging over the the whole 50ns or just using the last snapshot?
In the first case, the system might not be relaxed at first, you should leave some nano seconds of. In the second case you might be looking at a fluctuation.
Is the pressure in the system correct?

Groetnis,
Djurre

Please Log in or Create an account to join the conversation.

  • edroaldo
  • edroaldo's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 5 months ago #864 by edroaldo
Replied by edroaldo on topic Incorrect density
Dear Djurre,

Thank you for your replay!

I am using g_density as: g_density -f dyn.xtc -s dyn.tpr -n index.ndx -o denW.xvg. So, I averaging over the whole 50nm, right? The curve that I get for density profile looks like a "U" format. I have some regions above 1000kg/m³ and other regions below this value. I set up the pressure coupling with the parameters

Pcoupl = Berendsen
pcoupltype = anisotropic
tau_p = 5.0 5.0 5.0 0 0 0
compressibility = 4.5e-5 4.5e-5 4.5e-5 0 0 0
ref_p = 1.0 1.0 1.0 0 0 0

I look at the pressure with g_energy and looks like the pressure is wrong because the curve varies around zero. Any suggestion?

Thank you very much.

Please Log in or Create an account to join the conversation.

More
7 years 5 months ago #865 by djurre
Replied by djurre on topic Incorrect density
Dear Edroaldo,
The pressure might fluctuate, a lot actually, that normal, as long as the average is 1 bar.

I don't now how much the pressure is above 1000 kg/m3. If it is within a few percent, it is could be due to lack of sampling.

Please Log in or Create an account to join the conversation.

  • edroaldo
  • edroaldo's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 5 months ago #866 by edroaldo
Replied by edroaldo on topic Incorrect density
Hi, the average value of the pressure is 1.02bar, so the pressue is ok. The density is around 1150 kg/m³ in my system.

In relation to sampling, can you please tell me where I change this? I am using a configuration file similar to the MARTINI web site for DPPC membrane.

Thank you very much!

Please Log in or Create an account to join the conversation.

More
7 years 5 months ago #869 by djurre
Replied by djurre on topic Incorrect density
With sampling problem I mean you have to simulate (sample) longer. Simulate 1000 ns instead of 50.

Please Log in or Create an account to join the conversation.

More
6 years 4 months ago #1330 by skavyani
Replied by skavyani on topic Incorrect density
Dears,

I have this poroblem too.
Recently i simulated a system that contain a dendrimer molecule in a box of water. when i was trying to calculate density with g_density, the calculated density has a "U" shape but it's value is more than expected !.
After that, i simulate a single box with pure water but i gave the same result !!!.
the pressure average is 0.9999 bar and everything is fine except density !. i need correct density value because it can define my dendric system behavior.The spoilers are my .mdp and .top files for pure water system.

(NOTE:
-em tow times with two different methods.
-mdrun with 8 nodes.
)

.top
Warning: Spoiler! [ Click to expand ]

nvt :
Warning: Spoiler! [ Click to expand ]

npt :
Warning: Spoiler! [ Click to expand ]

md
Warning: Spoiler! [ Click to expand ]


I turned to CG martini force.f for less calculation time. unfortunately i can't do 1000ns simulation. this is so much. because best t.step that gave me a fine result with no system belowing up is 0.008ps for at least 50000 beads, with 8 nodes (more nodes make mdrun "DD" crash..i test many em setting and run methods ( as i can ) to use more nodes but i can't).


I would appreciate if you could help me to handle the problem.

Thank you very much.
Best regards,
S. Kavyani

Please Log in or Create an account to join the conversation.

More
6 years 4 months ago #1334 by djurre
Replied by djurre on topic Incorrect density
If you're simulating a box of water you should not need two EM steps and you do not need to do a NVT relaxation. Also if you can not simulated at a timestep longer then 0.008 ps, something is wrong.

When you generate the box of water, do you use genbox? Do you set -vdwd 0.21 ? If you open the initial box of water in, for example, VMD do you see any difference in density in a certain area? Genbox has the tendency to make two overlapping cubes inside large CG water systems.

Fix it by first making a much larger box using genconf and use that as input for genbox.

Please Log in or Create an account to join the conversation.

More
6 years 4 months ago #1335 by skavyani
Replied by skavyani on topic Incorrect density
Dear Djurre,

Many thanks for your reply.


For two em and nvt steps you are right. i was scrupulous!

I'm sorry, i was saying "same result" didn't mean that the pure water system has "U" shape. i just mean that the calculated density has a value more than expected.for pw sys density profile has a flat shape with but with wrong value, a value around 1200.
For miss understanding > i'm sorry. that's my fault

For p.w system i can use longer t.s but for dendric sys, i can't.
something is wrong with "W" beads.
why for such a simple system like pure water, density has such a ERROR?

I use "water.gro" that has been uploaded in matrini w.site, with 400 "w" beads (is for water.gro in martini-w.site -vdwd set to 0.21 ?! ) and then use genbox to fill larger box with this command:
genbox -cs water.gro -cp empty_box.gro -o water_box.gro -nmol 10000
NOTE:
Warning: Spoiler! [ Click to expand ]

in vmd the box is uniform. water_box and empty_box.gro are in size of 7 x 7 x 7 nm dim. the output has 3175 water beads in 7 7 7 box dim.

NOTE:
Warning: Spoiler! [ Click to expand ]


Best regards,
S. Kavyani

Please Log in or Create an account to join the conversation.

More
6 years 4 months ago #1337 by skavyani
Replied by skavyani on topic Incorrect density
Dears,

I apologize to anyone who wasted his time reading my first post.
I ignored that the tau_p parameter can affect my density profile. the value that has been used was 3.0 for "parrinello-rahman " and it better be around 0.5.

S. Kavyani

Please Log in or Create an account to join the conversation.

Time to create page: 0.109 seconds