rattle: too many constraint iterations.

You successfully prepared your system. Now: How do I tell Hippo to produce a trajectory from it? What parameters to choose? And what should I do when things go wrong.
Post Reply
kaihsu
Posts: 10
Joined: Wed Nov 12, 2008 3:23 pm
Contact:

rattle: too many constraint iterations.

Post by kaihsu »

I got this error when running a Monte Carlo simulation with implicit solvent and membranes:

rattle: too many constraint iterations.

Then Hippo quits.

What should I check? Is it because the protein is too large, or there are too many clashes, or something more insiduous?

Input file:

hippo input file

mode MD simulation
input type [PDB,zmx] PDB
input file kir21spm_hippo.pdb
max_steps 1e4
T [K] 100
use_pbc 0
cutoff [A] 44.0
cutoff_skin_depth [A] 6.0
freq_pdb_out 1e3
freq_xyz_out 1e3
freq_E_out 1e3
freq_state_restart_out 1e6
freq_check 1000
print_speed 1
conrot window_move_prob 0.25
conrot a 150
fixed chain ends 0
prob rigidbodymove/MC scan 0.0
max rigidbody move size [A] 0.2
use_gbsa 1
gb_mode 2
gb_mode2_printdetails 0
gb_cutoff [A] 8
united atom for aliphatic C 0
r_probe [A] 1.4
sigma 0.005
intra molecule epsilon 2.0
use_gb_flex123 (flexible 1-2 1-3) 1
use_gb_membrane 1
sigma_membrane -0.030
born_mode [1=exp 2=gauss] 2
sasa_mode [0,1=v(z) 2=gauss 3=exp] 2
v(z)_mode [1=linear 2=gauss 3=exp] 2
gamma_born -2.0
gamma_born2 -2.0
gamma_sasa -1.5
gamma_v(z) -2.0
L [A] 15
kaihsu
Posts: 10
Joined: Wed Nov 12, 2008 3:23 pm
Contact:

Re: rattle: too many constraint iterations.

Post by kaihsu »

Ah of course! I should have used ‘mode MC simulation’ rather than ‘mode MD simulation’. After this change, it works. (Thanks to Martin for the tip.)
starless
Posts: 3
Joined: Fri Sep 30, 2011 11:52 am

Re: rattle: too many constraint iterations.

Post by starless »

And what should I do if I get this error when I want to make a GBIM MD simulation? Without rattle it's rather slow (1 step/s).
martin
Developer
Posts: 38
Joined: Sun Jun 15, 2008 10:27 pm

Re: rattle: too many constraint iterations.

Post by martin »

Hi,

Sorry, the implicit membrane is currently only implemented in MC simulation mode. We've demonstrated that it is completely equivalent to MD simulations (it's actually a little faster in terms of sampling per CPU hour). Do you have a protein or a peptide?

Best wishes,
Martin
starless
Posts: 3
Joined: Fri Sep 30, 2011 11:52 am

Re: rattle: too many constraint iterations.

Post by starless »

Hi Martin,

Thank you for the reply. If GBIM is implemented only for MC, what will I get if I run MD with IM turned on?
martin wrote:Do you have a protein or a peptide?
I have a large protein with extracellular domain, helices parallel to membrane and TM helices. I know that TM helices are perpendicular to the membrane. Extracellular part is modelled based on good X-ray data, and membrane part is modelled based on cryo-EM data, my idea is to optimise membrane part and understand how deep in the membrane it sits to prepare it for explicit membrane simulation. Could you recommend me reasonable parameters for such run? My idea is to try different positions along z axis, but 1M MC steps (313 K, like in GBIM example with low probability of rotation) seems to be a bit not enough.
martin
Developer
Posts: 38
Joined: Sun Jun 15, 2008 10:27 pm

Re: rattle: too many constraint iterations.

Post by martin »

Hi,

You can do a transrot scan in the implicit membrane. The mode is "transrot scan"
And you can control the z, theta (tilt), and phi (rotation) angles by adding the following lines to your input file:
zmin (transrot scan) [A] -60
zmax [A] 60
zinf [A] 500
dz [A] 2
dtheta (tilt) [deg] 5
dphi (twist) [deg] 20

You might need to increase the sigma_membrane. Let me know if this helps!
Best wishes,
Martin
starless
Posts: 3
Joined: Fri Sep 30, 2011 11:52 am

Re: rattle: too many constraint iterations.

Post by starless »

Hi Martin,

It seems that transrot scan with default parameters is totally wrong: it gives the global minimum with z = -60 and helices parallel to the membrane. I don't really understand the difference between the tilt and twist, but I definitely know that my helices must be parallel to the membrane normal, and it seems that I need translational scan only. Which parameters should I use?

How does sigma_membrane affect the results? I used default value of -0.03, to what value should I change it?

Thank you in advance.
martin
Developer
Posts: 38
Joined: Sun Jun 15, 2008 10:27 pm

Re: rattle: too many constraint iterations.

Post by martin »

Hi,

Yes, sigma_membrane is the parameter that sets the hydrophobicity of the membrane, this is somewhat speculative and has to be increased for larger proteins (since it scales with the 'lipid accessible' surface area. If you increase it then at some point your protein should start inserting, -0.03 is too low, try -0.15. The tilt angle is the angle with respect to the membrane normal (theta). Phi is the twist around the axis of the protein (defined by the moment of inertia of the protein). Check the output pdbs for the minimal configuration. The bilayer is indicated by TST atoms at +-15 angstrom.
Good luck and let me know how you fare.
Cheers,
Martin

it should look like this:

mode transrot scan
...
use_gbsa 1
gb_mode 2
gb_mode2_printdetails 0
gb_cutoff [A] 8
united atom for aliphatic C 0
r_probe [A] 1.4
sigma 0.0049
intra molecule epsilon 2.0
use_gb_flex123 (flexible 1-2 1-3) 1
use_gb_membrane 1
sigma_membrane -0.1
born_mode [1=exp 2=gauss] 2
sasa_mode [0,1=v(z) 2=gauss 3=exp] 2
v(z)_mode [1=linear 2=gauss 3=exp] 2
gamma_born -3.0
gamma_born2 -3.0
gamma_sasa -1.5
gamma_v(z) -3.0
L [A] 15
zmin (transrot scan) [A] -60
zmax [A] 60
zinf [A] 500
dz [A] 1.0
dtheta (tilt) [deg] 10
dphi (twist) [deg] 45
Post Reply