[gmx-users] secondary structure element constraining

[alex rayevsky]

Dear users!

Did anybody meet the problem of positional constraints applied to the
secondary structure, namely keeping the 310-helix stable all the time? I've
modelled the substructure - a bundle of alpha helices with a 310 helix
segment (longer, thinner, reactive) on the one of them. And I need to hold
this conformation intact, to study other rearrangements. I'm not sure that
freezing is a good way to nail 3-5 residues in a huge system, maybe it is
somehow robust. At the same time I tried to simulate the helix with
restraints (10000 kJ/A^2 XYZ per each atom of the 310 helix part) and the
helix moved to the low energy state with a transition '310 -> alpha
conformation' without any obstacles. Usually I'm using COM pulling code,
but I don't know how to put some constraint on the backbone of these
several residues (forming triangle instead of the square, while looking
through the helix) without any relation to a surrounding system (without
reference atoms), just like a freezing but more realistic, allowing some
harmonic oscillations. Any suggestions??

Thank You!