Unsupervised hierarchical clustering of CNAs identifies distinct patient subgroups.

<p><b>A</b>) Unsupervised hierarchical clustering of raw log2 ratios derived from 72 serous type ovarian cancers. Copy number values are color coded as follows: blue (loss), white (normal) and magenta (gain). The pattern of dendrogram suggests two major genomic subgroups within the grade 3 tumors. <b>B</b>) PFS Kaplan-Meier plot for the two subgroups. <b>C</b>) Comparison of clinical characteristics between the patient subgroups. Histology: red = serous; Grade: orange = grade 2, yellow = grade 3; Stage: red = Ic, blue = II, green = IIc, yellow = IIIa, orange = IIIb, brown = IIIc, pink = IV, dark gray = IVa; Status: red = evidence of disease, blue = no evidence of disease; Outcome: green = complete remission, orange = progression, yellow = partial remission, brown = lost to follow up, pink = benign; 6 month progression: red = yes, blue = no, green = P (progression); Recurrence: brown = yes, orange = persistent disease, yellow = no; Platinum response: red = sensitive, black = resistant; Drug: blue = yes, light blue = no; Ascites: red = yes, black = no; Chemo: orange = yes, brown = no; Radiation: red = yes, black = no; General: white = n/a and/or blank.