Novel computational approaches to mutation discovery in tumour genomes: new tools to understand, diagnose and treat cancer

The Human Genome Project, which had the goal of sequencing the entire human genome, took more than 10 years, involved the work of thousands of people and cost more than $1 billion. Today, this same amount of work can be accomplished on a single machine in 10 days at a cost of $10,000, which halves every 18 months. The emergence of this "Next Generation Sequencing" (NGS) technology can reveal the precise genetic mutations that underlie how cancers develop, how they become more aggressive and how they acquire resistance to chemotherapy. A challenge of this technology is the data generated can be voluminous, complex and error prone; a single genome can produce over a terabyte of data.

Dr. Sohrab Shah is developing a new generation of computational tools using machine-learning approaches to improve accuracy and best interpret the large scale NGS data sets. With his clinically focused collaborators, he will then apply these technologies to sequence the tumour genomes from patients with triple negative breast cancer, ovarian clear cell carcinoma, and childhood osteosarcoma tumours — three cancer subtypes that do not respond to standard therapies. They hope to identify and profile unknown mutation patterns — or "mutation landscapes” — in each of these diseases.

These mutation landscapes will help Dr. Shah’s team further understand the biology of these tumours and provide a rational basis for the design of novel therapies to improve patient outcomes. His work will also include studying small populations of cancer cells to determine how they influence patients’ responses to treatment and how they become resistant to chemotherapy — two of the major issues facing oncologists today.