Project Spotlight: Phylogenetic Models for Predicting Cancer Progression

Center for Machine Learning and Health researcher Russell Schwartz, PhD, is a professor of Biological Sciences and Computational Biology at Carnegie Mellon University where he’s been involved with computational biology research since 2002. Dr. Schwartz and his research partner, Jian Ma, PhD, are working on a suite of software tools that clinicians will use to improve cancer diagnosis and therapeutic strategy based on the molecular signatures of the patient’s tumor genome.

Please share a little about your background and your research experiences.

I got into the field originally as an undergraduate researcher more than 25 years ago and have been working ever since to use the tools of computer science to solve problems in biomedical research. I did all of my studies at MIT before going to work at Celera Genomics, a startup company created to sequence the human genome. I joined Carnegie Mellon in 2002 and have had the opportunity to work in many areas of computational biology research, including aspects of biophysics, population genetics, genomics, systems biology, and evolutionary biology. My primary research focus for more than a decade has been computational cancer biology, which I first entered when my oncology collaborator, Dr. Stanley Shackney, and I realized the power that phylogenetics (evolutionary tree inference) could have for understanding how cancers develop and progress. The study of tumor evolution has since proven to be significantly important to our emerging understanding of cancer genomics and precision medicine and led to many unanticipated directions in my own lab including our Pittsburgh Health Data Alliance work.


What led you to the Pittsburgh Health Data Alliance?

I believe that academic research and industry, including entrepreneurship, have important and complementary roles to play in driving the scientific enterprise from fundamental research to practical improvements in the lives of individual people. While academic research is where I believe I can best contribute to the future of biology and medicine, I recognize that other contributions are needed to take that knowledge and put it into practice saving patient lives. The Alliance offered a unique opportunity to bring some of the research ideas we believe have translational potential out of the lab and into an environment where they can directly impact human health.


Walk us through your funded project.

Our project is designed to bring insights from basic research in cancer biology into clinical practice by using genomic data to improve our ability to predict how cancers develop, a process generically known as tumor progression. Much of the work builds off of past research of the Ma and Schwartz labs on developing computational technologies for basic cancer research. The Ma lab brings expertise in variant calling – the task of discovering the complicated ways cellular genomes are mutated in cancerous cell populations. The Schwartz lab brings expertise in tumor heterogeneity and phylogenetics, specifically the task of interpreting how an initially healthy group of cells evolves into a highly diverse population of cells carrying numerous genetic abnormalities. Our project is aimed at using machine learning inference to take the insights of how a tumor develops to predict how the tumor is likely to continue to evolve in the future. That general kind of prediction has relevance to numerous aspects of clinical cancer care, such as figuring out which precancerous lesions are likely to become cancers, which cancers have good or bad prognosis, or which of those with bad prognoses might respond durably to different therapies.


How is the PHDA uniquely positioned to assist your team and grow your project to commercialization?

Bringing new ideas in data-driven precision cancer medicine from the research laboratory to the clinic requires cross-disciplinary collaboration. It needs computational biology researchers developing new ideas and technologies, experimental biologists to inform and validate these methods, clinical collaborators who are knowledgeable about patient treatment but immersed in and appreciative of the role of basic research in driving translational advances, and business experts experienced in bringing these advances to real-world practice. The PHDA occupies a unique position connecting these and other key contributors across Carnegie Mellon University, the University of Pittsburgh, and UPMC.


What are your project’s next steps?

We have focused so far on bringing ideas from basic research to a demonstrable proof of practice in the research lab and are eager to work with the PHDA to carry them the next step forward into a viable product that can be brought from the lab to end users in clinical and industry practice.


What do you foresee the future of innovation looking like here in Pittsburgh?

Pittsburghers take pride in the city’s ability to bounce back from the economic troubles of the late twentieth century, in large part by reinventing itself through investment in healthcare and higher education. While this has led the city to become a center of innovation in many areas of science and technology, there is perhaps nowhere the city is more uniquely positioned for success than in the rise of computational and data-driven biomedical research and in the fields of genomics, systems biology, and precision medicine. I believe the PHDA can bring together these local strengths and play a major role in Pittsburgh’s rise as a leading international center of 21st century biomedicine.