Today, quantitative ideas are an essential part of cancer research. Discoveries made in the last decade, largely through high-throughput experiments generating large and complex data, have highlighted that most cancers are not homogeneous diseases, but rather, collections of different diseases, each with its own biology, and each requiring different therapeutic approaches. Many of the opportunities and challenges for the cancer research community involve using these complex data to understand the key features of each individual cancer subtype, and with this information identify or develop the appropriate therapy for each patient.
The Department of Biostatistics and Computational Biology (B&CB) contributes to this mission in two ways: 1) by designing innovative tools that keep the Institute's research at the forefront in its ability to handle data challenges and 2) by providing access to highly qualified and committed faculty members to teams of investigators across the Institute, for which quantitative sciences are essential.
To the first goal, faculty in the Department conduct research in statistical, mathematical and computational methods that are directly helpful for cancer research. These methods are integrated into clinical and translational research, population-based studies, and cancer biology. Faculty develop software for research and clinical applications and perform high-throughput genomic experiments guided by their computational insights.
To the second goal, all members of the faculty collaborate extensively in research initiatives of other investigators, providing expert advice on experimental design, data collection, data storage, data integration and data analysis.
Members of the Department are trained in the art of learning from complex experimental data. Their participation with the Institute's research activity is not only a necessity, but a mechanism to boost the speed, efficiency and quality with which the data generated by the Institute's research is translated into usable knowledge.
Currently, the Department of Biostatistics and Computational Biology is comprised of 32 faculty, 6 research associates, 18 research fellows, 18 MA biostatisticians and 14 bioinformaticians. In addition, there are at least 23 graduate students working with faculty members at any time. During 2012, Department members authored or coauthored more than 240 peer-reviewed publications.
Department members play central roles in the Dana-Farber/Harvard Cancer Center (DF/HCC), the largest NCI-designated Comprehensive Cancer Center in the country. Giovanni Parmigiani, PhD, is the DF/HCC Associate Director for Population Sciences and the leader of the Biostatistics and Computational Biology program at DF/HCC. The DF/HCC Biostatistics Core is located at Dana-Farber and is directed by faculty member Paul Catalano, ScD. The Core provides consultation and assistance to Cancer Center members in all DF/HCC research programs. Members of the department are involved in a variety of DF/HCC activities, and play a role in the development of all clinical research protocols by serving as members of DF/HCC's Scientific Review Committee as well as the Institutional Review Board.
The Department of Biostatistics and Computational Biology has historically been at the forefront of the development of innovative approaches for carrying out clinical trials of cancer treatments. Its founder, Marvin Zelen, PhD, now emeritus, is one of the most influential innovators in this area. Today, because of the increasing personalization of cancer therapy, and because of the importance of therapies that target features of cancers that are not shared by all patients, traditional clinical trial paradigms are insufficient to guide us to the best drugs in a timely fashion. Several faculty (Barry, Gelber, Gray, Parmigiani, Trippa) are active in developing innovation in clinical trial design and analysis to meet this challenge.
Also, as the result of its rich history, the Department is home to the statistical coordinating centers of two important consortia, the Eastern Cooperative Oncology Group, or ECOG, (Gray) and the International Breast Cancer Study Group, or IBCSG, (Gelber, Regan). These groups support the design and analysis of trials conducted simultaneously in multiple centers nationally and internationally. Through these consortia, our faculty plays a leadership role in the design and conduct of clinical research globally. These centers are also a platform for progress in the science of clinical trials.
More than 90% of potential new cancer drugs tested in clinical trials never find their way to pharmacy shelves. The dearth of effective new cancer drugs is impeding the ability to deliver personalized cancer medicine — the right drug for the right patient at the right time. Traditional clinical trials accrue patients based on the anatomical locations of their tumors. However, cancer is a disease in which certain genes undergo changes (mutations) that result in cancer. Each type of cancer has several subtypes and the genetic mutations may vary in each subtype, which means that two patients with lung cancer or breast cancer or any other type of cancer may not benefit from the same treatment. It follows that a genetically-based clinical trial, in which patients are divided into subgroups based on the specific genetic mutations driving their tumors, has the potential to have a greater impact than a traditional clinical trial. The application of sophisticated bioinformatics and computational algorithms to voluminous genomics data makes it possible to design and conduct genetically-based trials, and recruit the right patients for the right clinical trials. This type of trial design may in fact increase success rates in bringing new FDA-approved cancer therapies to the clinic.
Health care decisions in cancer need to be supported by evidence of the effectiveness, benefits, and harms of different prevention and treatment options. This evidence is generated from research studies that compare drugs, medical devices, tests, surgeries, or other ways to deliver better cancer care or to prevent cancer. While clinical trials are an essential component of this process, we need to utilize a variety of other sources of information. Massive data sets are collected as part of routine clinical practice, and as part of the financial management of health care.
Epidemiological studies that follow large populations to monitor their health contain extensive information about cancer and its causes. Imaging and sensing technologies generate high-resolution information about patients on a routine basis. All of these are examples of data sources that are potentially very rich, but also substantially more challenging to understand, as compared with clinical trials. A mission of the Biostatistics and Computational Biology Department is to provide the logical and computational underpinning for deriving solid conclusions from the investigation of these data sources.
Patho-epidemiology is the emerging science of integrating pathology data and tumor biomarkers at the RNA, DNA and protein level into population-based epidemiological studies. It poses important novel challenges to which the Department is actively contributing (Parmigiani, Tyekucheva), in collaboration with the Center for Molecular Oncologic Pathology (CMOP) and the Harvard cohorts. As an example, B&CB faculty are developing computational approaches that can take data from a sample that contains a mixture of cancer and non-cancerous adjacent tissue, and separate the information from each. This allows us to study the interaction between a tumor and its surrounding tissue and to investigate why some tumors remain in place while others spread across one's body.
The Department is home to a thriving community of investigators working on advanced approaches to cancer research using data generated by high-throughput molecular biology assays, such as microarrays and next-generation sequencing. This research covers the development of analytical methods, software, and data infrastructures. Areas of application include somatic mutation analysis (Michor, Parmigiani, Trippa), expression analysis for clinical applications (Barry, Parmigiani, Quackenbush), epigenomics (Irizarry, Liu, Michor, Yuan), transcriptional regulation modeling (Liu, Yuan), next-generation sequencing analysis (Irizarry, Liu, Quackenbush), metabolomics, and others.
Established in 2008, the Center for Cancer Computational Biology (CCCB) is one of Dana-Farber's Integrative Research Centers and a vital component of the Institute's Strategic Plan for Research to advance the era of personalized medicine. The CCCB provides broad-based support for the analysis and interpretation of genomic and other large-scale data, furthering basic, clinical, and translational research by providing new ways of understanding human cancer. The Center is developing new methods for improving the analysis and interpretation of genomic data through the integration of diverse data types with the goal of creating open-source software tools to be made freely available to the research community.
One of several major accomplishments has been the design and implementation by John Quackenbush, PhD, of an integrated clinical and research data warehouse. This sophisticated tool enables researchers to merge information from multiple sources and make it readily accessible, reducing human error and bypassing problems inherent in linking together different types of data. The CCCB provides Dana-Farber investigators with state-of-the-art assistance in the collection, management, analysis, and interpretation of large-scale data and it provides software, services, and training to assist investigators in advancing their personal research.
The Center for Functional Cancer Epigenetics, co-led by Myles Brown, MD, and Shirley Liu, PhD, explores the key role that epigenetic alterations and abnormal transcriptional regulation play in the development and progression of cancer. A better understanding of these alterations will lead to better diagnosis for cancer and the potential to contribute to the knowledge required for the development of new therapeutics involving epigenetic mechanisms.
Cancer cells arise as the result of the accumulation of genetic changes that make these cells function differently from their intended purpose. Genetically altered cells survive and sometimes thrive in human tissues through a process that is very similar to evolution and natural selection. Mathematical models of evolution are very powerful tools for understanding how this process works and are pursued actively by Department faculty (Michor, Parmigiani).
For example, cancer cells often harbor hundreds of small changes (mutations) in their DNA code. While some of these (the "driver" mutations) result in cancer occurrence, many others (referred to as "passenger" mutations) represent neutral variation that does not affect how cancer develops. The identification of "driver" mutations is of crucial importance for drug discovery because they represent promising targets for therapeutic intervention. Michor, Parmigiani and Trippa have all contributed computational, statistical and mathematical tools that helped sharply define this concept. Their work has helped to identify mutations that act as drivers during tumorigenesis with the potential to aid in the prioritization of candidate mutations for functional validation that will contribute to the process of drug discovery.
The Department of Biostatistics and Computational Biology is home to the Physical Science – Oncology Center (PS-OC), directed by Franziska Michor, PhD. The principal mission of the PS-OC is to advance our understanding of the physical principles that govern cancer initiation, progression, response to treatment, and the emergence of resistance. The members of the PS-OC include theoretical biologists from Dana-Farber Cancer Institute, Memorial Sloan-Kettering Cancer Center, and City College of New York, and scientists from Vanderbilt University and Memorial Sloan-Kettering Cancer Center. Collaborations between theoretical biologists and experimental scientists in the PS-OC will bridge the divide between the physical sciences and oncology.
The Department has a close partnership with the Department of Biostatistics at Harvard School of Public Health, where the majority of B&CB faculty have primary academic appointments. Department faculty are leaders of training grants and curriculum development initiatives, direct research, teach in the doctoral program, and also teach in the undergraduate degree program in statistics at Harvard College.
Three department faculty members received significant awards in the last year. Marvin Zelen, PhD, founding chair of the Department, received the American Cancer Society's Medal of Honor, the highest honor bestowed by the ACS in recognition of outstanding contributions to cancer control in three categories. Richard Gelber, PhD, was a co-recipient of the Brinker Award, one of the most prestigious awards given for breast cancer research and supported by Susan G. Komen for the Cure. The White House honored John Quackenbush, PhD, as an Open Science Champion of Change his contributions to making large and complex sets of biological data widely accessible to the larger research community.
Search by last name:
Browse directory of
Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA 02215 | Call us toll-free: