NeAT - RNSC

RNSC - Restricted Neighbourhood Search Cluster Algorithm
The RNSC program was developed by Andrew King.
RNSC is an efficient cost-based local search clustering algorithm that explores the solution space to minimize a cost function, calculated according to the numbers of intracluster and inter-cluster edges. (King, 2004, M.Sc. thesis; King et al, 2004)
The stand-alone version of RNSC is available upon request.
   Input format 

Comment on the demonstration example

This demonstration graph consists in the yeast co-immunopreciptation interaction dataset described in Gavin et al (2006). It contains 1430 nodes and 6531 edges. The rnsc algorithm is applied on it in order to highlight clusters of densely connected polypeptides.
Graph

Upload graph from file :

   
Column specification (only relevant for tab-delimited input)
Source node
Target node
Maximum number of cluster
Tabu length
Tabu list tolerance
Naive stopping tolerance
Scaled stopping tolerance
Diversification frequency
Shuffling diversification length
Number of experiments