AW: Coarse-grained reverse transformation

From: Norman Geist (
Date: Mon May 13 2013 - 00:43:00 CDT

Hi Fugui,


I never did coarse grained simulations and such converting, but if you think
about it, you will notice, that it is impossible to convert a coarse grained
structure to a perfect all-atom structure again. Think how you would do it
by hand, and notice the questions you would ask yourself. Such a converting
takes place, while replacing the coarse residue balls, with the original
all-atom residues. Therefore one would at least need information about the
orientation of side chains and the backbone torsions, for example in
proteins. But the coarse grained structure doesn't provide all that required
information, also like steric relationships between side chains etc. If
your structure looks otherwise ok, you could try to carefully minimize and
equilibrate the new structure to the all-atom environment, maybe while
securing H-bonds or other parts of the molecule or slow heating to keep the
new conformation, or to prove it's actuality.


If I'm wrong, feel free to correct me.


Good luck.


Norman Geist.


Von: [] Im Auftrag
von Fugui
Gesendet: Samstag, 11. Mai 2013 17:21
Betreff: namd-l: Coarse-grained reverse transformation


Dear NAMD users,

I tried to converse a coarsed grained protein structure into an all-atom
protein structure by using VMD. Firstly, i used the coarsed grained
structure and the all-atom structure which is original structure before the
MD simulation to get a starting structure which is shown in the attached pdf
file. There are several long C-C bond which can cause problem in the
annealing simulation in the next step. Did anyone meet the problem before?

Any suggestion is appreciated.

Best wishes



This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:11 CST