[MMTK] Question about minimization

Matias Saavedra matias_saavedra at yahoo.com
Thu Jan 5 21:16:06 CET 2006

Hi Konrad. 

I have been running a minimization with a protein with
223 residues. I used Steepest Descent for the first
steps until a Gradient Norm of 50, and Conjugate
Gradient for the rest. The work has been running for
48 hours in my power book G4 1,33 GHz (each
minimization step takes around 1 minute), and the
Gradient Norm doesnt go under 2,0. For the last 24
hours of calculations, the Gradient Norm has not
changed very much, ranging between 7,0 and 1,0. Now
the process is in the step number 3500 with a Gradient
Norm of 3,68.

What does it mean? Do i have to run more time the
calculations until i reach the value of 0,01 for the
Gradient Norm or, simply, the method can't find a

What can i do ?


Yahoo! DSL – Something to write home about. 
Just $16.99/mo. or less. 

More information about the mmtk mailing list