Systematic Identification of Driver Networks in Cancer
A vast number of mutations contribute to cancer, but the observed non-random combinations of those leading to transformation highlight the importance of hallmark pathways and networks in cancer progression. While many pathways have been implicated in cancer, attributes such as tumor heterogeneity, tissue of origin, and degree of progression lead to each case exhibiting a unique subset of altered pathways. Taken together, this diversity among cancer types and their origins has complicated the development of targeted cancer treatments. We propose here to systematically identify the protein networks driving cancer, across a range of tumor types starting with head and neck squamous cell carcinoma and breast cancer. Coupled with functional validation and high-resolution structural analysis of the key protein interactions and complexes, we anticipate major insights into the underlying tumor biology as well as the potential to unravel genetic vulnerabilities of therapeutic relevance.

Mapping the Pharmacogenetic Landscape for Precision Medicine
It is well known that cancer is tremendously heterogeneous with few tumors having the same set of mutated, amplified, or deleted genes. Clearly these molecular differences alter a tumor’s responsiveness to chemotherapy, but current knowledge of how the tumor genotype influences drug sensitivity is poor. We will seek to vastly increase our understanding of pharmacogenetic interactions in cancer (gene-gene and gene-drug interactions). Recognizing that oncogenic transformation requires alteration of the function of many genes, we will use state-of-the-art high-throughput epistasis mapping and data analysis pipelines to systematically interrogate the function and pairwise interactions of a panel cancer driver genes and therapeutic targets in both head and neck squamous cell carcinoma and breast cancer, expecting to identify many new synthetic lethal relationships. Anticipating the discovery of multiple therapeutically relevant synthetic lethal interactions, we have already formulated a plan for rapid clinical testing of the most promising hits as new treatment arms on the I-SPY 2 trial in breast cancer. Through this work, we expect to develop fundamental new insights into the genetic logic and functional synergies underlying cancer pathways as well as to greatly expand the ability of clinicians to practice precision oncology.

Using Networks to Seed Hierarchical Whole-Cell Models of Cancer
Knowledge of cell biology is often modeled in the form of molecular networks and interaction maps, consisting of sets of genes and gene-gene (or protein-protein) pairwise interactions. In reality, however, biological systems are not simply one large protein network, but consist of a deep and dynamic hierarchy of functional subsystems ranging across many orders of magnitude in scale. Here, we move beyond basic interaction maps to instead use molecular interaction data to develop hierarchical structure/function models of the cancer cell. This hierarchical structure will be developed using the protein-protein interaction data generated here and backstopped by public networks; it will provide an objective definition of a cancer cell by systematically identifying the hierarchical relations among its associated systems of genes and proteins. We will next use this descriptive hierarchy to seed a predictive whole-cell model of cancer. This hierarchical model will be validated and revised by applying it to predict therapeutic responses in PDXs of head and neck and breast tumors as well as inform an ongoing I-SPY 2 breast cancer clinical trial.