Skip to main content

Table 1 Overview of algorithms tested

From: A systematic comparison of genome-scale clustering algorithms

    Allows Overlapping Clusters
     Pre-specified Number of Clusters ( k )
      Thresholded Correlations
Method Type Result Range     Parameters Tested
Ward Hierarchical k   Y   Average cluster size
Average Hierarchical k   Y   Average cluster size
McQuitty Hierarchical k   Y   Average cluster size
Complete Hierarchical k   Y   Average cluster size
k-Means Partitioning k   Y   Number of clusters
SOM Neural network k   Y   Grid size/typea
QT Clust Partitioning 24-385     Maximum cluster diameters
CAST Graph-based 1-6162    Y Threshold
CLICK Graph-based 4-32     Cluster homogeneity
WGCNA Graph-based 4-160     Power, Module detection method
NNN Graph-based 23-52 Yb    Minimum neighborhood size
k-Cliques Communities Graph-based 1-68 Y   Y Threshold, Clique size
Maximal Clique Graph-based 1,000-64,000 Y   Y Threshold
Paraclique Graph-based 8-615 Yc   Y Threshold, Glom factor
  1. Clustering methods are listed by name, along with the type of algorithm, and a general listing of parameters tested. Number of clusters in the result, given the parameters and data set tested, is only provided here as an approximate figure. Empty results are obviously not included. aGrid type can be either rectangular or hexagonal, in an m x n layout. We tested both types, but used an m x m layout for simplicity (k = m2). bRarely occurs in practice. On this data set we observed no overlap with NNN. cOptional, not used in this analysis.