PPI-GA: A Novel Clustering Algorithm to Identify Protein Complexes within Protein-Protein Interaction Networks Using Genetic Algorithm

Table 11

Cliff’s effect size test.

Algorithm	Recall	Precision	F-measure
DMCL-EHO	.0 (neg.)	.5 (large)	.5 (large)
K-means	.5 (large)	.0 (neg.)	.5 (large)
MCODE	.5 (large)	.5 (large)	.5 (large)
MCL	.5 (large)	1 (large)	.5 (large)
ClusterOne	.5 (large)	1 (large)	.5 (large)
Ramadan GA spectral-DC	.5 (large)	.5 (large)	.5 (large)
Ramadan GA spectral-MC	.0 (neg.)	1 (large)	.5 (large)
Ramadan GA spectral-NC	.0 (neg.)	1 (large)	.5 (large)
Ramadan GA spectral-RC	.0 (neg.)	1 (large)	.5 (large)
Ramadan GA random-DC	.5 (large)	.5 (large)	.5 (large)
Ramadan GA random-MC	.0 (neg.)	1 (large)	.5 (large)
Ramadan GA random-NC	.25 (small)	.5 (large)	.5 (large)
Ramadan GA random-RC	.5 (large)	.5 (large)	.5 (large)

A positive value indicates the effect size favor of the PPI-GA. The interpretation of the effect size is indicated in parenthesis; neg. stands for negligible; med. stands for medium.