From: mason H (
Date: Wed Aug 27 2014 - 21:38:53 CDT

   I have an enzyme that I am trying to perform dynamical network analysis
on. I am following the protocol described in the Dynamical Network Analysis
Tutorial pdf file (downloaded from NAMD/VMD website).

   I have a question on the “restrictions” section of the “networkSetup”
input file (aka parameter file). In the tutorial example, each protein
residue is defined as a node. The node is 'represented' by the CA atom of
the residue. They implement the “notNeighboringCAlpha” restriction. This
preclude node edges between neighboring CA atoms.
    My question is this: Why exactly would you want to implement this
restriction? Suppose you want to look at the number of suboptimal paths
between two distant nodes. Why could a path not traverse along, for
instance backbone atoms of a group of adjacently numbered residues? – Why
would you disallow such a path?

Now, to complicate matters...
  Suppose I set up a system using 'clusters,' whereby each protein residue
is divided into two nodes; Backbone and Side-chain. Backbone nodes are
'represented' by CA atoms , and side nodes are 'represented' by terminal
side chain atoms (like NZ for lysine).
     If I was to follow their example, I would apply the
“notNeighboringResidue” restriction and, I believe, the "notSameResidue"
restriction. (I believe adding the notNeighboringCAlpha would be redundant
with notNeighboringResidue). But why would I want to add any of these
restrictions? Couldn't a path traverse along the side chain node of a
residue, through to the backbone node of that same residue, and on to the
backbone of the adjacent residue?

   Lastly: Would it make any sense in my clustered example to keep the
“notNeighboringResidue” restriction, but discard the “notSameResidue”
restriction [thereby enabling connections to form between backbone and
side-chain nodes of the same residue, but disallowing any edges between
neighboring residues] ?

Thank you