Reference Clustering method Classification method Cancer Clinico-biological significance
Golub TR et al.,1999 [4] Self- organizing maps Linear discriminant analysis ALL vs. AML A distinction that is critical for successful treatment
Alizadeh AA et al., 2000 [5] clustering Hierarchical N/A DLBCL Subtypes with differences in the stages ofB-cell differentiation (germinal  center,activated B-cell) and their relation withprognosis
Perou CM et al.,2000; Sorlie T etal., 2003; van 'tVeer LJ et al.,2002; van deVijver MJ, et al. 2009 [8,9] Hierarchical clustering Various: PAM,correlation withcategories Breastcancer Subtypes (basal, luminal A, luminal B,HER2+) with differences with regard toclinical features, response to treatment,and prognosis.
Salazar R et al.,2011 [12]   A nearest meanclassifier CRC Classifier that can predict patients withearly-stage CRC
Collisson EA et al.,
2011 [28]
NMF N/A PDA Subtypes of PDA and their responses totherapy
Sadanandam A etal., 2013 [29] NMF PAM CRC Subtypes of CRC and their relation withcellular phenotype and response to therapy
Armstrong SA, etal, 2003 [40] N/A K-nearest-neighbors MLL Rearrangement of the MLL genedistinguishes a unique leukemia
Tan IB, et al 2011 [11] NMF SVM GC Subtypes of GC and their relation withsurvival and response to chemotherapy
Budinska et al,2013 [29] Hierarchicalclustering Multiclass LDA CRC Subtypes with molecular heterogeneity
Table 1: Some machine learning methods used in cancer subtyping.