Biotechnology, statistics, collaboration, and the pursuit of the future

Canada has embraced the potential for biotechnology to revolutionize modern life. Substantial investments are being made in biotechnology in a range of fields including health care, the environment, agriculture, aquaculture, and sustainable development. Biotechnology is a strong contributor to the Canadian innovation economy and Canada places highly in international rankings of biotechnology development. However, capitalizing on investments in biotechnology is challenging due to the complexity of the biosystems being created. Biotechnology depends critically on analyzing massive streams of complex data arising from new technologies to gain knowledge, make predictions, and evaluate risk and uncertainty. These data analysis problems arise in all areas of biotechnology, e.g., from gene sequencing to GPS data to chemical properties to health outcomes, and at every step of the biotechnology process from initial investigation to end-product evaluation. Biotechnology data analysis problems require development of new statistical methodologies. They cannot be solved using brute-force computational methods and traditional analytics. Indeed, using those approaches plays a substantial role in the growing crisis of non-reproducible research, whereby findings cannot be replicated or are not robust with respect to realistic operating conditions. Negative consequences of non-reproducibility include raising questions about the value of biotechnology and wasted investment. The solution to the challenges inherent to biotechnology data analysis is to foster interdisciplinary collaborations between statistical scientists and biotechnology researchers that lead to the creation of cross-disciplinary scientific innovations. The success of such an approach is demonstrated by the Canadian Statistical Sciences Institute (CANSSI). Headquartered at Simon Fraser University, CANSSI is a national institute supported primarily by the Natural Sciences and Engineering Research Council. CANSSI is Canada’s catalyst for discovery and innovation in statistical sciences and application of statistical sciences in other fields. CANSSI programs support the pursuit of cutting-edge collaborative research involving statistical sciences along with the communication and application of the results to science, engineering, and society. For example, the Collaborative Research Teams (CRT) program fosters statistical sciences research and training interactions that span disciplines and institutions. The CRT program emphasizes the co-creation of knowledge, the leveraging of Canada’s expertise in statistical sciences, and the synergy achieved by embedding partnership into interdisciplinary projects at their inception. CRT projects have initiated significant advances in research across a swath of applications and have had a direct impact on Canadian society in several areas. Some CRT projects tackle the development and application of statistical methodology important to biotechnology, including: Analyzing whole-genome sequencing to better predict drug resistance, phylogenetic or epidemiological relatedness, and identifying genomic variants. Using massive data streams related to the movement, behavior and health of humans and animals to make better health predictions. Combining biological and fisheries data into models to assess the condition of stock and predict the response to varying levels of fishing pressures. Determining and describing familial dependence structures in DNA sequence data for the investigation of rare genetic mutations involved in complex diseases. Quantifying and mapping severe fire risk, with a focus on wildland-urban interface. The success of the CANSSI model shows that interdisciplinary partnerships between statistical scientists and biotechnology researchers provide the best approach to tackle the complex data analysis problems endemic to biotechnology. Unfortunately, CANSSI is the sole Canadian national institute in statistical sciences with the mandate to interface across the entire spectrum of science and engineering. CANSSI simply does not have sufficient resources to fulfill the interdisciplinary capacity of the Canadian statistical science community with respect to biotechnology. Finding the resources to foster more interdisciplinary collaborations with statistical sciences is essential for capitalizing on Canada’s investment in biotechnology. Donald Estep is the director of the Canadian Statistical Sciences Institute and a Canada Research Chair in Computational Probability and Uncertainty Quantification at Simon Fraser University in Burnaby, British Columbia. He has served on scientific advisory panels for the U.S. National Science Foundation and Department of Energy and on the Sandia National Laboratories CISE External Review Board and has co-authored several reports. The Hill Times