Volume 5, Issue 1, Pages 216-223 (October 2013) Signaling Network Assessment of Mutations and Copy Number Variations Predict Breast Cancer Subtype-Specific Drug Targets  Naif Zaman, Lei Li, Maria Luz Jaramillo, Zhanpeng Sun, Chabane Tibiche, Myriam Banville, Catherine Collins, Mark Trifiro, Miltiadis Paliouras, Andre Nantel, Maureen O’Connor-McCourt, Edwin Wang  Cell Reports  Volume 5, Issue 1, Pages 216-223 (October 2013) DOI: 10.1016/j.celrep.2013.08.028 Copyright © 2013 The Authors Terms and Conditions

Cell Reports 2013 5, 216-223DOI: (10.1016/j.celrep.2013.08.028) Copyright © 2013 The Authors Terms and Conditions

Figure 1 Analysis of Integrated Networks for Breast Cancer Cell Survival and Proliferation The data of genome sequencing, genome-wide RNAi screening, copy number variations, and gene expression profiles of individual lines were used for constructing an integrated network for each individual cell line. Cell-line-specific networks across each of the breast cancer subtypes were used for constructing subtype-specific networks for cancer cell survival and proliferation. Comparative and differential analysis of the subtype-specific networks allowed us to predict subtype-specific treatments and significantly classify breast tumor samples. See also Figure S1 and Table S1. Cell Reports 2013 5, 216-223DOI: (10.1016/j.celrep.2013.08.028) Copyright © 2013 The Authors Terms and Conditions

Figure 2 Hierarchical Clustering of the 16 Breast Cancer Cell Lines Hierarchical clustering of the cell lines using cell-line-specific network hubs: (A) driving-regulator hubs, (B) essential gene hubs, and (C) the hubs of essential genes and driving regulators combined. Red and beige in the heatmaps indicate whether the hub genes are present or absent, respectively, in a cell line. See also Table S3. Cell Reports 2013 5, 216-223DOI: (10.1016/j.celrep.2013.08.028) Copyright © 2013 The Authors Terms and Conditions

Figure 3 Subtype-Specific Survival Signaling Networks Subtype-specific survival signaling networks for basal A (A), basal B (B), and luminal (C) subtypes. Nodes represent genes while links represent regulation (directed links) or interaction (neutral links) between genes. A node is represented by a pie chart that shows each gene’s distribution as essential gene (red), a driving-regulator (blue), or a proliferation-influencing gene (cream) in its subtype. The background color behind the clusters represents a cluster’s function in relation to one of the cancer hallmarks: apoptosis (pink), cell proliferation (green), and cell cycle (blue). Cytoscape (Saito et al., 2012) was used to present and visualize the networks. See also Figure S2 and Table S4. Cell Reports 2013 5, 216-223DOI: (10.1016/j.celrep.2013.08.028) Copyright © 2013 The Authors Terms and Conditions

Figure 4 Clustering of 16 Breast Cancer Cell Lines and 402 Breast Tumor Samples Using the Hubs from Subtype-Specific Networks (A) Hierarchical clustering of the 16 cell lines using the differential hubs from the subtype-specific networks of luminal and basal subtypes. In the heatmap, for a given cell line, a hub gene appears in red if it is an essential gene, a driving regulator, or a proliferation-influencing gene; otherwise, it appears in beige. On the side bar, gray and yellow represent luminal and basal cell lines, respectively. (B) The same differential hubs from (A) were used to classify 402 breast tumor samples. In the heatmap, red represents mutated genes or amplified genes that are among the top 50% of the expressed genes for tumor samples. Cell Reports 2013 5, 216-223DOI: (10.1016/j.celrep.2013.08.028) Copyright © 2013 The Authors Terms and Conditions

Figure S1 Data Sources and Cutoff Values for Defining Driving Regulators and Essential and Proliferating-Influencing Genes, Related to Figure 1 For a given cell line, if a gene is among Top 75% of the expressed genes and its RNAi-screening p value < 0.05, it is defined as an essential gene in that cell line; if a gene is among Top 75% of the expressed genes and its RNAi-screening 0.05 > p value < 0.1, it is defined as a proliferation influencing gene in that cell line; if a gene is driver-mutating gene, or has a GISTIC score > 0.3, plus among Top 50% of the expressed genes, and if its RNAi-screening p value < 0.4, it is defined as a driving-regulator for cell survival in that cell line. Cell Reports 2013 5, 216-223DOI: (10.1016/j.celrep.2013.08.028) Copyright © 2013 The Authors Terms and Conditions

Figure S2 Dartboard Showing Overlap of Essential Genes and Driving Regulators for Cell-Line-Specific Networks of the Subtypes, Related to Figure 3 Overlapping of essential genes for basal A (A), basal B (C) and luminal (E) and overlapping of driving-regulators for basal A (B), basal B (D) and luminal (F). The outer most circle colored with peach (basal A and basal B) or gray (luminal) represents the number of genes that do not overlap with any other cell line within their respective subtype, and are therefore unique to that cell line. Going toward the center, the bull’s eye contains the number of genes shared by all the cell line-specific networks in that subtype. Cell Reports 2013 5, 216-223DOI: (10.1016/j.celrep.2013.08.028) Copyright © 2013 The Authors Terms and Conditions