sbeG = mcl(g, e, r, n, a, iter) 
 
This function implements Markov Clustering Algorithm to deduce 
appropriate number of non-overlapping clusters. 
 
g    - Current network 
       =============== 
       The current loaded network. 
 
 
e    - Expansion coefficient (power) 
       ============================= 
       The positive power coefficient to which the matrix must be raised 
       before the inflation step (higher power gives fewer clusters, runs 
       slower for large-scale networks). 
       Default: 2 
 
r    - Inflation coefficient 
       ============================= 
       The positive value to which each matrix column must be raised to 
       make the matrix, column stochastic. 
       Default: 2 
 
n    - Minimum value for matrix pruning 
       ================================ 
       Minimum value for pruning. All elements less than this value 
       will be made 0. 
 
a    - Allow each node to connect to itself 
       ==================================== 
       Before iteration, allow self edge connection for each node 
       within the network (Enabling self connections for larger networks 
       improves the speed of the algorithm). 
       Default: False 
 
iter - Number of iterations the algorithm should run 
       ============================================= 
       The algorithm runs for this many number of iterations and quits. 
       It will also quit if it is able to deduce clusters before approching 
       it's limits. 
       Default: 27 
 
Ref: Stijn van Dongen. A cluster algorithm for graphs. 
     Technical Report INS-R0010, 
     National Research Institute for Mathematics and Computer Science in 
     the Netherlands, Amsterdam, May 2000. 
 

mcode, clusteronerun 
 
Systems Biology and Evolution Toolbox (SBEToolbox). 
Authors: Kranti Konganti, James Cai. 
(C) Texas A&M University. 
 
$LastChangedDate: 2012-07-26 12:13:07 -0500 (Thu, 26 Jul 2012) $ 
$LastChangedRevision: 221 $ 
$LastChangedBy: konganti $