By Franklin Faust
Ellen Wijsman will admit that she doesn’t have all the answers, but she might just have all the right questions. As Professor of Biostatistics in the UW School of Public Health and UW Division of Medical Genetics/Department of Medicine in the School of Medicine, she has a healthy skepticism for new scientific discoveries about Alzheimer’s disease. If something doesn’t add up during a science talk—whether it’s an unsubstantiated assumption, a subtle methods glitch that complicates the interpretation of the science, or a study population too small to detect a robust correlation between a hypothetical risk factor and disease—Wijsman speaks up, asks questions, and ultimately improves understanding by exposing logical flaws and offering solutions. The moment her voice fills the room at a presentation, ADRC researchers turn in their seats, anticipating an incisive reality check.
Wijsman focuses on using statistical analysis for defining the inheritance patterns of Alzheimer’s disease and identifying connections between a risk factor and disease in distinct population groups. In other words, she uses mathematical models to find meaningful patterns and insights into disease risk hidden within massive amounts of medical data.
In the lab, as in meetings, her reputation for scientific rigor and high standards proceeds her. “In our lab, she is famous for saying, 'Nice result—now can you make it go away?' She always encourages us to double check assumptions and think through their implications,” says Nicola Chapman, PhD, a research scientist in Wijsman’s laboratory who has worked with her for 25 years, “She has definitely nurtured in me a deep skepticism of surprising results, which in the end results in better science.” By repeatedly holding herself and others to a gold standard, Wijsman brings out the best in those around her.
She first arrived at the UW ADRC in 1987 to fill the role of liaison between statistics and genetics on an ambitious project to identify genes involved in familial Alzheimer’s disease. For 15 years, Wijsman has served as Lead of the ADRC’s Data Management and Statistics Core (DMS), directing data support for many ADRC projects and providing assistance with statistical analysis to junior researchers. Her retirement from this position marks the end of 33 years of direct involvement with our Center. As she moves on to a role of collaboration and consultation, she leaves behind a large pair of boots to fill. Her work as a mentor, teacher, and scientist has advanced scientific discovery and elevated the careers of those around her in a way that continues to help our Center thrive.
A Knack for Study Design
The scientific integrity of a study, and credibility of its findings, starts with a good plan. A study design is a research road map that plans out how the data generated for a study can be effectively analyzed and interpreted. Without a solid framework, researchers cannot answer the questions about links between genes, health, and disease that they were inspired by in the first place.
Wijsman's passion for refining study designs has helped a countless number of researchers to course-correct their plans for successful interpretation of data. Whether it’s her real-time feedback at scientific presentations, fielding of investigators’ project pitches as the manager of the ADRC’s DMS Core, or feedback on grant proposals made to the many review panels she has served on, such as the Center for Inherited Disease Research Access Committee, Wijsman has consistently given her colleagues strong feedback when she sees a better path to answering their research questions. “She’ll tell you your approach to the study is wrong, and she’ll tell you why it’s not appropriate. She’ll be critical, and then she’ll give you alternatives,” says Elizabeth Blue, PhD, Associate Professor of UW Medical Genetics, and one of Wijsman’s longtime collaborators. “She’s not telling you not to do it. She’s telling you how to do it better. And she’s usually right. She’s always got her eye on the ball.”
In our lab, she is famous for saying, 'Nice result—now can you make it go away?' She always encourages us to double check assumptions and think through their implications.Nicola Chapman, PhD
Wijsman’s many years of criticism and constructive feedback have had a tremendous impact on the field, one research plan at a time.
In fact, the ADRC’s first genetics project benefited from Wijsman’s eye for good study design, as it led to a new recognition of the genetic complexity of familial Alzheimer’s disease. She took a critical eye to the Center’s familial Alzheimer’s disease dataset and found differences in the patterns of inheritance from parent to child, particularly in the ages at which people developed dementia. Wijsman intuited that the search for a single Alzheimer’s gene was overly simplistic. She suspected that varying patterns of genetics and biology seen in different populations all converge on the similar syndrome of Alzheimer’s-type dementia, a phenomenon termed genetic heterogeneity.
She urged the team to use age-of-onset data to group families into subsets, which made it possible to detect the genetic signals they were looking for and find the distinct genetic variants that promote Alzheimer’s disease in individual families. In that first project, Wijsman set the course for the Center’s early achievements in identifying single gene causes of Alzheimer’s disease and frontotemporal degeneration. "Wijsman was indispensable to the breakthroughs in the genetics of Mendelian Alzheimer’s disease that established us as a successful Center,” says Thomas Grabowski, MD, ADRC Director.
Over thirty years later, Wijsman now plays an important role in sorting out the genetic puzzle of late-onset sporadic Alzheimer’s disease through her work in the Alzheimer’s Disease Sequencing Project (ADSP), a massive scientific effort to tease out the genetic factors influencing the more common form of Alzheimer’s disease. Researchers think that the patterns of disease inheritance in late-onset Alzheimer’s disease family pedigrees are more complicated than for early-onset because the development of the late-onset disease may be influenced by multiple genetic variants and environmental factors. To have a chance at understanding this complex pattern of genes and disease, statistical analyses require a larger sample size–more genetic data from more people. One problem is that the task of searching for complex genetic relationships in big populations with massive amounts of genomic data requires long computer processing times, making it more difficult to carry out the analyses. Wijsman has been a leader in trying to make these analyses work.
“Ellen is developing novel solutions to real world problems that the whole field is going to have,” says Blue. “Her guiding question is: ‘How can we make a resource for other people so that they don’t have to spend 2 years doing the work that we had to?’." While building statistical tools to analyze the ADSP’s massive and complex genetic data sets, Wijsman found a blind spot. There was a gap between the data collected and the analyses that needed to be performed to find genetic variants that increase disease risk. “We couldn’t see the differences in the genetic variants between the Alzheimer’s cases and the controls,” says Blue. “But Dr. Wijsman’s work shifted the way they handled the follow-up study so that we could have the data we needed to move forward.” Wijsman’s diligence led to the ADSP’s addition of a “Discovery Extension Phase” that gives researchers the data they need to progress in the search for genetic variants that increase the risk for late-onset Alzheimer’s disease. Over more than thirty years of research on both early-onset and late-onset forms of Alzheimer's disease, Wijsman’s laser-focus on identifying problems in research and addressing them with new methods and ideas has pushed the field forward time and again.
Developing some of those statistical analysis tools took decades of dedication, largely stemming from the struggles Wijsman faced in her early years at the ADRC. “The data analysis was very challenging in those early years,” she says. “The challenge drove my interest in the development of a methodology that would eventually allow computations to be done in practical amounts of time.” Wijsman collaborated with Elizabeth Thompson, PhD, Professor Emerita in the UW Statistics department, and Adjunct Professor Emerita in UW Genome Sciences and UW Biostatistics, to develop new statistical tools that could better model complex genetic traits and make analysis more manageable. Some of their first projects leveraged the heart disease research family pedigrees of Arno Motulsky, MD, a founder of the field of medical genetics whose work originally attracted Wijsman to move to the UW.
Over the next 27 years, through waves of new genetic and computational technologies, Thompson and Wijsman continued to build new tools, such as software and lines of code that have given researchers more power. Their innovations allow for detection of genotyping errors in pedigrees, statistically sound assignment of genetic background to individuals who lack data in large pedigrees, and an improved ability to use data from multiple points in the DNA at once, to name a few outcomes. With the tools to leverage new genetic technologies, Wijsman and Thompson’s colleagues have been able to investigate questions about complex interactions of genetic and environmental risk factors in human diseases that were previously out of reach.
“Dr. Wijsman embraced the interface of statistics and neurogenetics, and the particulars of neurodegenerative disease. She is an inspiring example of how enabling a true interdisciplinary commitment can be,” says Grabowski.
Through the career-long collaboration, Wijsman linked Thompson’s statistical expertise to the expertise of medical geneticists and neurogenetics researchers at the ADRC. “Working with Ellen let me bring statistical models and computational methods to address real biological problems,” says Thompson. The two collaborators have unraveled some of the mysteries behind Alzheimer’s disease, heart disease, prostate cancer, and alcoholism, finding new genetic and environmental factors that may increase the risk of these conditions.
A Mentor’s Dedication to Training the Next Generation
When Wijsman arrived at the University of Washington, the institution did not have a structure in place to train new statistical geneticists. “What was lacking was training of the next generation of statistical geneticists,” says Thompson. To help ensure the future of their field, Wijsman and Thompson started a training track of statistical genetics courses in UW Biostatistics and Statistics Departments. They worked for decades to instruct new generations of researchers. “Her training in quantitative and population genetics, and also her great ability and enthusiasm for it, connected people on the medical side to the broader statistical genetics world. She continually strives to bring those worlds together,” says Thompson. At a scientific crossroads, Wijsman has provided statisticians a bridge to explore medical questions and has helped medical geneticists use mathematical models to draw conclusions from real-life data about disease risk. People trained and instructed by Wijsman have gone on to use statistical analyses to find genetic variants and epidemiological risk factors that influence susceptibilities to a wide variety of human diseases.
Wijsman’s past students remember her capstone course in statistical genetics as a fundamental part of their training, without which their education would not be complete. “Ellen taught her students the value of being a statistician in solving real scientific problems. I think Ellen must have been a role model for so many of them,” says Thompson. France Gagnon, MSc, PhD, Research Professor and Associate Dean at University of Toronto’s Dalla Lana School of Public Health, whose research leverages modern molecular technologies and analytic approaches to identify genetic influencers of heart disease, recalls the enthusiasm that Wijsman brought to the classroom. “She spends more time preparing for class than most people in her class probably spend learning the material. She is passionate about statistical genetics and it shows in the way she teaches,” says Gagnon. Wijsman’s teaching record shows her commitment to lifting up and challenging others in her field to improve.
Wijsman’s passion for statistical genetics and skill in teaching its fundamentals and application to medical questions shined in the laboratory as much as the classroom. “Ellen was really my most influential mentor that I've had in my career,” says Weiva Sieh, MD, PhD, Associate Professor in the Departments of Population Health Science & Policy and Genetics & Genomic Sciences at the Icahn School of Medicine at Mount Sinai. Under Wijsman's guidance, Sieh worked to search for the genetic players in a fatal dementia complex that affects the Chamorro people of the island of Guam as well as other Pacific Island populations. “She was such a good role model for scientific integrity as well as personal integrity. Some of the most lasting lessons that I learned about how to conduct myself as a scientist, what it means to be a mentor and a researcher, how to balance having a child and maintaining a high-level career, I learned from just modeling Ellen,” says Sieh.
As Wijsman asked hard questions of her lab mates, she kept an open door for trainees to ask their own. “I felt like nothing was off limits and that she would be willing to answer my questions honestly,” says Sieh. Wijsman also consistently participates in a trainee-mentor lunch with the International Genetic Epidemiology Society (IGES), a society in which Wijsman served as President in 2005, where trainees come to get an outsider perspective on their questions or anxieties about the field and their careers.
The effect of Wijsman’s exemplary training did not stop with the researchers she trained, for those people have internalized her approach to mentorship. Gagnon largely credits her emulation of Wijsman’s teaching style for the Graduate Teaching Award that she received from the University of Toronto in 2014. “My trainees commented that they felt comfortable, that they were able to ask lots of questions, and that I helped them grow as an independent scientist,” she says. “When I saw that, I said, 'Wow, that's Ellen Wijsman there.' "Wijsman’s personal guidance fostered her mentees’ curiosities and helped them grow into their lab coats, and they now carry that same spirit in working with their own scientists-in-training.
Wijsman equipped her trainees with the tools to collaborate with those who work in applied medicine, and they have continued to do so. “Back when I was training with Ellen, there were very few people in the statistics field collaborating so genuinely with clinicians and wet lab scientists,” says Gagnon. “I carry her influence into my work today, pushing for new research projects in my role, and working very closely with wet lab scientists, clinicians, and clinical psychologists to make sure we are thinking outside of the box when talking about solving problems.” Wijsman’s spirit of collaboration and willingness to learn and lean on the expertise of others persists in those she has trained.
As Wijsman prepares to retire from the ADRC DMS core, she looks forward to continuing to consult on various ADRC projects and share her perspective in scientific meetings. But her legacy speaks for itself, in generations of capable researchers who apply statistical tools to new medical questions. With their memories of Wijsman as a guide, these researchers will continue to advance our understanding of disease risk in the wider population and train the leaders of tomorrow. •
Cover photo: Shutterstock