A functional gene module identification algorithm in gene expression data based on genetic algorithm and gene ontology

Table 2 The comparative results of different clustering algorithms (DBSCAN, Affinity Propagation, Hierarchical Clustering, Spectral clustering, PAM, and PAM-GA) on BRCA, THCA, HNSC, COVID-19, Stem, and Radiation datasets

Datasets	Algorithm	Silhouette Score	Parameter Value
BRCA	DBSCAN	0.010419	min_samples = 3, eps = 0.6
	Affinity Propagation	-0.470702
	Hierarchical Clustering	0.097262	n_clusters = 8, linkage = ”average”
	Spectral Clustering	-0.331269	n_clusters = 8, gamma = 1E-04
	PAM	0.109636	k = 11
	PAM-GA	0.132069	k = 11
THCA	DBSCAN	0.037134	min_samples = 9, eps = 0.3
	Affinity Propagation	-0.558596
	Hierarchical Clustering	0.137183	n_clusters = 8, linkage = ”average”
	Spectral Clustering	-0.381021	n_clusters = 8, gamma = 1E-05
	PAM	0.129793	k = 10
	PAM-GA	0.178543	k = 10
HNSC	DBSCAN	0.036295	min_samples = 11, eps = 0.5
	Affinity Propagation	-0.474879
	Hierarchical Clustering	0.076271	n_clusters = 8, linkage = ”average”
	Spectral Clustering	-0.293960	n_clusters = 8, gamma = 1E-07
	PAM	0.109274	k = 8
	PAM-GA	0.132306	k = 8
COVID-19	DBSCAN	0.064597	min_samples = 13, eps = 0.2
	Affinity Propagation	-0.593570
	Hierarchical Clustering	0.128473	n_clusters = 18, linkage = ”average”
	Spectral Clustering	-0.431040	n_clusters = 8, gamma = 1E-06
	PAM	0.177675	k = 9
	PAM-GA	0.216090	k = 9
Stem	DBSCAN	0.161855	min_samples = 12, eps = 0.2
	Affinity Propagation	-0.698527
	Hierarchical Clustering	0.167391	n_clusters = 9, linkage = ”average”
	Spectral Clustering	-0.525442	n_clusters = 8, gamma = 1E-03
	PAM	0.239311	k = 9
	PAM-GA	0.280704	k = 9
Radiation	DBSCAN	0.023345	min_samples = 4, eps = 0.3
	Affinity Propagation	-0.564850
	Hierarchical Clustering	0.080373	n_clusters = 8, linkage = ”average”
	Spectral Clustering	-0.372413	n_clusters = 8, gamma = 1E-06
	PAM	0.092876	k = 9
	PAM-GA	0.142994	k = 9

^*The last column denotes the optimal parameter values obtained after many experiments

ISSN: 1471-2164