[gmx-users] Fwd: Selecting the temperature distribution

[Mark Abraham]

Likewise, I agreed with what Massimo said.

As an example, I recently did a fairly large set of REMD simulations of a
320-atom disordered peptide with rather more water and many fewer replicas
than you propose. I did so because I expected low barriers and large
maximum diameter (the latter from an implicit-solvent REMD). Even after the
fact, I cannot demonstrate that my decisions were correct (and probably
could not even if I could have produced clearly converged ensembles).

Mark