- Posts: 5
Too many LINCS warning during production run
- SNEHA
-
Topic Author
- Offline
- Fresh Boarder
Less
More
6 years 3 months ago #7469
by SNEHA
Too many LINCS warning during production run was created by SNEHA
Hi,
I am trying to simulate a system with protein inserted into a lipid droplet. A part of C-terminal of protein is manually inserted into a lipid droplet and to remove atomic clashes, the respective lipids are manually removed from the droplet. Both the lipid droplet and protein are well equilibrated before forming the above mentioned complex. This complex was energy minimized in vacuum as well as after solvation with standard MARTINI water using steepest descent algorithm. There were 2-3 LINCS errors while minimization. The atoms mentioned in the errors were either related to protein beads or lipids but none of them were present in the area where protein is manually inserted. Minimization was completed with a note that
Steepest Descents converged to machine precision in 22179 steps,
but did not reach the requested Fmax < 10.
I performed equilibration using Berendsen thermostat at 310K for 10ns at 2fs timestep and then 50 ns each for timesteps of 10fs, 15fs and 20fs. During this, position restrains were gradually decreased from 1000 to 1.
This system was subjected to production run without any position restrains. However, the simulation is terminated after 200ns with error "Too many LINCS warnings" creating many step.pdbs.
The total size of system is 199140 beads. Thus, I could not save each frame (nstxout =1) as it would generate huge files.
Kindly help me out in this regard.
ps - I could not get how to upload image of system :p
I am trying to simulate a system with protein inserted into a lipid droplet. A part of C-terminal of protein is manually inserted into a lipid droplet and to remove atomic clashes, the respective lipids are manually removed from the droplet. Both the lipid droplet and protein are well equilibrated before forming the above mentioned complex. This complex was energy minimized in vacuum as well as after solvation with standard MARTINI water using steepest descent algorithm. There were 2-3 LINCS errors while minimization. The atoms mentioned in the errors were either related to protein beads or lipids but none of them were present in the area where protein is manually inserted. Minimization was completed with a note that
Steepest Descents converged to machine precision in 22179 steps,
but did not reach the requested Fmax < 10.
I performed equilibration using Berendsen thermostat at 310K for 10ns at 2fs timestep and then 50 ns each for timesteps of 10fs, 15fs and 20fs. During this, position restrains were gradually decreased from 1000 to 1.
This system was subjected to production run without any position restrains. However, the simulation is terminated after 200ns with error "Too many LINCS warnings" creating many step.pdbs.
The total size of system is 199140 beads. Thus, I could not save each frame (nstxout =1) as it would generate huge files.
Kindly help me out in this regard.
ps - I could not get how to upload image of system :p
Please Log in or Create an account to join the conversation.
- peterkroon
-
- Offline
- Gold Boarder
Less
More
- Posts: 210
6 years 3 months ago #7483
by peterkroon
Replied by peterkroon on topic Too many LINCS warning during production run
Hi SNEHA,
how did you generate your protein topology and starting conformation?
And although I'm not a lipid expert, I think you may need more than 160 ns of equilibration.
In your production run, what barostat and timestep do you use? Try using the berendens barostat and a larger tau-p;
how did you generate your protein topology and starting conformation?
And although I'm not a lipid expert, I think you may need more than 160 ns of equilibration.
In your production run, what barostat and timestep do you use? Try using the berendens barostat and a larger tau-p;
Please Log in or Create an account to join the conversation.
Time to create page: 0.088 seconds