[gmx-users] Hessian instability

[Ignacio Rodriguez Fernandez]

Hello,

I´m trying perform a normal mode analysis of a large molecule. My problem is
that I always get negative eigenvalues, which shouldn´t happen in theory.
The problem seems to be the instability of the Hessian matrix, due to an
unproper minimization of the molecule. I have already tried some
minimization algorithms (steepest descent, conjugate gradients, l-bfgs,
etc)for GROMACS but they all seem to fail.

Do you know the real cause of this problem and how could be avoided?


Thanks,

Ignacio