BMC Bioinformatics

Table 3 Number of significantly enriched GO terms (or pathways)

From: Comprehensive evaluation of matrix factorization methods for the analysis of DNA microarray gene expression data

(a) Leukemia dataset
☐	Kmeans	SVD	PCA	ICA	NMF	SNMF	BSNMF
ALL	480	389	441	453	532	425	535
AML	85	262	223	222	167	266	280
Total	565	651	664	675	699	691	815
(b) Medulloblastoma dataset
☐	Kmeans	SVD	PCA	ICA	NMF	SNMF	BSNMF
classic	517	373	467	479	388	456	599
desmoplastic	58	361	226	213	335	208	206
Total	575	734	693	692	723	664	805
(c) Fibroblast dataset
☐	kmeans	SVD	PCA	ICA	NMF	SNMF	BSNMF
cluster1	52	45	71	47	57	41	128
cluster2	32	35	68	27	54	48	69
cluster3	48	24	63	61	37	75	50
cluster4	126	38	37	96	108	60	155
cluster5	54	63	60	33	65	68	102
Total	312	205	299	264	321	292	504
(d) Mouse dataset
☐	kmeans	SVD	PCA	ICA	NMF	SNMF	BSNMF
cluster1	593	520	294	258	637	686	690
cluster2	27	61	107	114	38	28	56
Total	620	581	401	372	675	714	746

Number of significantly enriched terms at α=0.05

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com