The Emerging Potential for Network Analysis to Inform Precision Cancer Medicine

Abstract

Precision cancer medicine promises to tailor clinical decisions to patients using genomic information. Indeed, successes of drugs targeting genetic alterations in tumors, such as imatinib that targets BCR-ABL in chronic myelogenous leukemia, have demonstrated the power of this approach. However, biological systems are complex, and patients may differ not only by the specific genetic alterations in their tumor, but also by more subtle interactions among such alterations. Systems biology and more specifically, network analysis, provides a framework for advancing precision medicine beyond clinical actionability of individual mutations. Here we discuss applications of network analysis to study tumor biology, early methods for N-of-1 tumor genome analysis, and the path for such tools to the clinic.

Keywords: cancer systems biology; network analysis; precision cancer medicine.

Figure 1

Overview of network applications in cancer.

Figure 2

Schematic diagram of three concepts used for identification of cancer drivers. Nodes represent proteins while edges represent interactions between proteins. A) Identification of driver pathways, where the nodes within the shaded region belongs to the pathway. B) Identification of driver genes, where red nodes correspond to the proteins encoded by driver genes within the pathway (outlined). C) Identification of driver mutations targeting the driver genes. Sizes of the nodes represent mutation frequency. Pie charts on the nodes display the percentage of driver (red) and passenger (pink) mutations.

Figure 3

Summary of several widely-used algorithmic approaches in the identification of driver pathways/subnetworks. A) An example network with nodes representing proteins and edges representing interactions between proteins. Nodes marked with an asterisk are mutated proteins. B) A heat diffusion process where each node is initially assigned a color based on the mutation score of the corresponding gene. Heat diffuses across the edges of the network where intensity of the node colors denotes the mutational influence on the protein. In the end, significantly mutated subnetworks (outlined) are reported. C) A prize-collecting Steiner tree approach where node size represents mutation score and edge width represents the confidence in the interaction (other concepts can be used for determining the node size and edge width). Algorithm reports the connected components with the most “prize” nodes and the least number of edges. D) A mutual exclusivity approach that identifies sets of genes that simultaneously maximize mutual exclusivity and coverage of patient samples.

Figure 4

A network-based approach for integration of protein-protein interaction networks with protein 3D structures and mutation data to identify cancer driver genes. Nodes represent proteins and edges represent interactions between proteins in the network. Unexpected mutational enrichment (red lollipops) on the interaction interface region of a protein (blue) implicates the encoding gene as potential cancer driver.

Figure 5

Network of network-based methods used in cancer. Nodes represent the name of the methods and edges represent similarity in the utilized algorithmic approaches. The colors of nodes represent the application of the method. Similar applications are grouped together (dashed line) under four main categories: driver identification for pathways, genes (or miRNAs or alternative splicing) or mutations, patient stratification, network biomarkers (therapeutic biomarkers, variant interpretation, subtype biomarkers, survival analysis, prognostic biomarkers) and N-of-1 analysis including transcriptomic analysis and driver prediction.