Good morning,
I was hoping to ask a specific technical question branching from an
earlier thread. Recently, Dr. Cheatham commented:
"It is becoming more and more difficult to publish the results of single
simulations since questions on reproducibility and convergence arise so
most practitioners tend to run multiple independent MD simulations with
different initial conditions (ion distribution, structure, velocities,
...) with a minima of three replicates."
I was hoping to ask a bit more about the ion distribution aspect. In
certain publications, I have read that ~100ns of equilibration is
necessary to achieve equilibration of the ion positions in a DNA
simulation (meaning the counter-ions one inserts), because the movement
of the ions is slow. Taken in conjunction with the above statement, this
would imply that one would want to do ~3 distinct equilibrations, not
just ~3 distinct trajectories? Obviously "3" is a placeholder here for
whatever number is genuinely appropriate based on convergence behavior.
I am accustomed to doing multiple "production" runs with various initial
conditions, multiple initial minimize/heat/equilibrate conformers, etc.,
but never multiple distinct equilibrations of ~100ns PER conformer
tested. Do I understand this properly?
--
Dr. Robert Molt Jr.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Oct 11 2021 - 07:30:03 PDT
