[gmx-users] LINCS vs SHAKE!

[Arman M. Soufiani]

Dear Gromacs users,

I want to run an NVT md. I am using LINCS algorithm to constrain both a
polymer and protein in my solvated system. 
Some of my dihedral types are tabulated functions. I am running the
equilibration on one node and there is no domain decomposition. 
All together, when I run the NVT my system explodes and I receive the
following warning and errors:
 *WARNING: For the 15 non-zero entries for table 0 in table_d1.xvg the
forces deviate on average 281% from minus the numerical derivative of the
potential*
*relative constraint deviation after LINCS:
rms 2.431943, max 26.861526 (between atoms 713 and 714)
bonds that rotated more than 30 degrees:*

I really have no idea what should I do?
I am opt for providing more information of the system I am working with.

Thank you in advance,

BR

Arman



--
View this message in context: http://gromacs.5086.n6.nabble.com/LINCS-vs-SHAKE-tp5002281.html
Sent from the GROMACS Users Forum mailing list archive at Nabble.com.