[gmx-users] Simulating a peptide

[Mark Abraham]

> Dear all,
>       I want to simulate a peptide of 10 residues length small part of a
>       protein of my interest), to see its secondary structure variation
>       upon mutation. Is it enough to simulate that 10 residues  alone to
>       see the change or is it mandatory to put entire protein and to
>       seek for posible way to constraint other uninterested region.

Probably the electrostatic environment of a small isolated peptide bears
no resemblance to that of the same peptide as part of a protein in its
normal biochemical milieu. Then, one has to decide whether there's any
reasonable way to cap the termini of the peptide. Then one needs to
justify the assumption that the phase space sampled by the peptide
resembles the phase space sampled in the protein... the latter is likely
to have strong geometric constraints imposed by the remainder of the
protein structure. You could model this in a couple of ways in an MD
simulation, but I wouldn't fancy defending any of them to a journal
referee.

Mark