CDS Distinguished Visiting Professor Charles DeLisi Reflects on 25 Years of the Bioinformatics Program
Charles DeLisi, dean emeritus of the Boston University (BU) College of Engineering, established the nation’s first PhD program in bioinformatics at BU in 1999. Since then, the program has grown by leaps and bounds, producing notable alumni such as Rhonda Harrison, who became the first Black person and first woman in the United States to earn a bioinformatics PhD. In 2023, the Bioinformatics Program at BU moved into the Faculty of Computing & Data Sciences (CDS). The move was made to encourage and facilitate research collaboration between computational biology and CDS’s wide range of computational and data-driven fields. In this Q&A, DeLisi, a CDS Distinguished Visiting Professor, offers his thoughts on how the program has evolved since its inception, how he came up with the idea for the program, how ethics fit into bioinformatics, and more.
How has the Bioinformatics Program evolved from 1999, when it was established, until now?
When we started the program, it was the only game in town. We were able to recruit superb students, and we also had a strong outreach program; for example, we gave presentations at Historically Black Universities. But the question of whether we would be able to maintain our lead and momentum even as other universities began emulating us was very much on our minds.
We did, in fact, continue to attract superb students — and in spades. I can name a number of stellar recent Ph.D. dissertations, but I’ll avoid the temptation of giving a list because I don't want to chance leaving out something important.
I will, however, name a particular dissertation because, in addition to being excellent research, it’s relevant in a number of other ways. I'm referring to the 2024 dissertation by Jacquelyn Turcinovic, Transmission Dynamics and Rare Clustered Transmission Within an Urban University Population Before Widespread Vaccination. She worked with John Connor and won the award bearing my name, and I'm, of course, honored to have that association. It is notable that their paper is far from an isolated instance of infectious disease research here at BU. Strikingly, BU President Emeritus Dr. Robert Brown wrote an outstanding paper, A simple model for control of COVID-19 infections on an urban campus, which was published in Proceedings of the National Academy of Sciences (PNAS). A still different approach was taken by my colleague Gyan Bhanot and myself in our model of the dynamics of COVID-19 spreading across the European continent.
All this is to say that there is considerable interest here at BU in infectious disease dynamics, as there would be in any highly interdisciplinary area of inquiry relevant to multiple species and backed by extensive data warehouses — characteristics that fit perfectly with CDS.
How do you think the program is doing in terms of recruitment of minority students?
The backstory is interesting. About 25 years ago — I can't believe it's that long — I received a telephone call from Nancy Koppell, one of BU's most distinguished faculty members, who said a senior National Science Foundation (NSF) official was visiting and wanted to speak with me. He told me that NSF was about to establish something called the Integrated Graduate Education and Research Training program, and he encouraged an application. Nancy and I immediately started tossing ideas around. In the end, BU was in the first small cohort of recipients, and the award was large enough to enable the start of our program.
One of the NSF requirements for a successful application was strong minority outreach — and their philosophy aligned perfectly with our own. It is no accident that in the first several years, three African American women received their PhDs here, one of whom was the first African American female in the nation to receive a PhD in bioinformatics. That was Rhonda Harrison, who did her dissertation in my lab and coincidentally will be here at the end of September to receive the College of Engineering distinguished alumni award. One of my interests, now that I have some free time, is to establish our program in bioinformatics and systems biology as the national leader in recruitment of underrepresented minorities—or at least to help stabilize its lead if it's already on top.
Incidentally, you might be surprised at the reference to Bioinformatics and Systems Biology (BSB), rather than just Bioinformatics. Some 20 or so years ago, I don't remember the exact year, the faculty voted unanimously to change the name to include systems since it was apparent even then that there was a growing emphasis on what I'll loosely call genetic networks. In fact, our annual international workshops with Japan and Germany included systems biology in the title, but for some reason, the name never took hold here at BU.
BU BSB was the first such program in the nation — what made you think of it?
In the early part of my career, mathematics and computation played a very limited role in the biological and medical sciences. It was outside the mainstream and almost impossible to have a career in biomedical research without a wet lab. But that all changed abruptly with the start of the Human Genome Project (HGP). Although mathematically knowledgeable scientists from a number of disciplines started to look at the challenges that would be posed by an HGP data flood, the computer science community was especially important. It's fair to say that without computer science, there would be no genomics, or proteomics, or metabolomics — or whatever your favorite “omics.”
Getting the HGP off the ground was a real battle, but as it gained support, a number of biologists began to recognize that cell biology was rapidly becoming an information science. I happened to be one of them, and since I was a mathematical/computational biologist by profession and also involved in HGP policy, it is not surprising that computer science and mathematical education was on my mind. What is surprising is that nothing happened between the time I left the Department of Energy (DOE) in the late ’80s and the time I received the call from Nancy, about a decade later. Such is the inertia of education and research to disruptive initiatives. And I'm almost embarrassed to say that I was not immune to inertia, being caught up in only decanal matters.
The genomic revolution also had its downsides, did it not?
The ethical dilemmas that it raised, including civil liberties and genetic selection, were recognized from the outset. A few months before leaving DOE, I set aside 3% of the budget for ethical and legal studies. Six months later, Jim Watson arrived at the National Institutes of Health (NIH) and took similar action. I was very hopeful that the substantial funding commitments made by DOE and NIH would make a difference, but they didn't — it took two decades before the first piece of legislation was passed protecting privacy — and I would call it a “quasi de minimis” bill. Ethics was not a high priority in Congress. In fact, I would say that the problem of not seeing it as an urgent matter goes beyond Congressional apathy and is in part traceable to the way education is organized.
What do you have in mind?
Let me give you some additional examples, and then we can see what they might have in common. I thought it was very important to include a course on ethics as a core requirement for a bioinformatics degree, but the idea encountered considerable faculty pushback. It was included only after I made the case that excluding it wouldn’t go over well with NSF and would likely impact the chance of continued funding. Scientists are focused on science, and ethics too often is just a distortion from another culture.
How was the course greeted by students?
Extremely well. It was taught by Michael Baram, a law professor with an undergraduate degree in engineering. The class was split roughly equally between law students and BSB students, and they all had diverse undergraduate backgrounds, so multidisciplinarity was present from the outset. It was very exciting, even more so because every year, two or three term papers by BSB students were accepted for publication in a refereed journal (BioLaw and Ethics).
I can add one more example related to ethics. In the late ’90s, the chair of sociology and I co-taught a new course, Technology and Society. The course was heavily oversubscribed and drew students from across the University. Ten years later, I co-taught it again with a first-rate sociology faculty member who was up for tenure. The CAS Dean was not enthusiastic about her, and she wound up moving to UMass-Amherst as a tenured professor and founding chair of a center for science and society.
Our sociology department fortunately had another very well-qualified faculty member, Ruha Benjamin, who happened to be an African American female. It didn't take long to realize that she was an absolute superstar and that we were probably in danger of losing her. So, about halfway through the semester, I spoke to her chair about my concerns. She was sympathetic, but the dean wasn’t. Two weeks before the end of the spring semester, Ruha told me she had an offer from Princeton and was planning to leave. So much for the course.
Each of these three examples, in my opinion, illustrates a lack of familiarity with an important aspect of culture and learning by otherwise highly educated individuals. So I again go back to the idea of the importance of space that houses multiple cultures, breaking down the spatial silos that accompany our departmental structures without disrupting departmental cohesion.
Your interests seem well aligned with the mission of CDS. Can you elaborate a bit more on multidisciplinarity?
Yes, but perhaps I should begin by distinguishing between interdisciplinary and multidisciplinary. The importance of the former, i.e., of research in which investigators trained in different fields work toward a common goal, has long been recognized. One of the interesting aspects of CDS is that it not only enables interdisciplinary research, but it has a multidisciplinary faculty. A mix of faculty from many disparate fields not obviously related but occupying common ground provides the seed corn for interdisciplinary projects that might not now be apparent but which could develop if interactions are frequent enough.
Can you give me an example?
Around 15 years ago, I couldn't have imagined that the research being carried out on transcriptional regulation by Mark Kon and myself could have had implications for financial portfolio analysis, and then by chance, I met a faculty member from MIT's mathematical finance program and discovered that our machine learning methods were precisely what was needed to construct profitable, low-volatility portfolios.
That's very interesting. Do you think that financial experts working in proximity to computer scientists could help foster interactions and perhaps lead to new approaches?
The concept that unrelated disciplines will start talking to one another if they share common ground is very important and goes well beyond AI and finance, which is now old hat. I’m reminded of two pieces of sociology that bear on this. One was the description by C.P. Snow, the English chemist and novelist, of the two cultures at British universities, which he said — if I recall correctly — were separated by a gulf of mutual incomprehension and sometimes hostility and dislike. The related story is the finding by the groundbreaking sociologist Robert K. Merton that prejudice in interracial housing projects is significantly less than it is in nonintegrated communities.
And so it is in academic communities. Proximity leads to understanding, without which intellectual bridges cannot form. It's important to keep in mind that academic departments are human creations; nature can care less about them. But aside from these general remarks about multidisciplinarity, there's a more personal interest for me. Mark, whom I continue to work with, has an office in the data science building — as does my long-term colleague Nancy Kopell. And then there are the young BSB faculty with whom I share numerous biological interests.
There's also a strong orthogonal appeal — the energy source for the building: It's entirely geothermal. As I discussed in my memoir, The Mysterious Commonplace, I've had an interest, albeit discontinuous, in environmental issues for most of my life. My current focus on climate change started around five years ago when I read an article in the New York Times Sunday Magazine proposing that land management could be an important part of the answer. I happen to agree with the philosophy, but as I indicated in a letter to the Times, the article was lacking in one very important respect — it failed to even mention the most powerful set of tools available for managing land: agricultural genomics. And I suggested that the reason for the failure was that the two communities knew next to nothing about one another.
I followed by publishing a simple model illustrating the possibilities and then, with support from the Sloan Foundation, organized an international conference that brought together researchers from various communities and made a first pass at scoping out the research landscape. The findings were summarized in a paper published in BioDesign Research, a journal that had recently been started by the American Association for the Advancement of Science (AAAS). That was followed by our search for novel and effective methods for drawing down atmospheric carbon and related approaches to reducing atmospheric nitrous oxide and methane.
Our focus was almost perfectly complementary to the goals of Bill Gates's Breakthrough Energy Ventures Inc., which focuses on fossil fuel alternatives, so they supported us very generously for five years. In early 2024, my colleagues and I edited a book published by Cold Spring Harbor Press on synthetic and systems biology approaches to climate change mitigation and adaptation. There’s a lot of interest here at BU in climate change, much of it complementary to my own, and I can't imagine that in the kind of atmosphere provided by the data science building, different threads won't soon converge.
Do you have any other goals?
Taking a deep dive into the role of artificial intelligence (AI) in international relations (IR), which until now has had essentially no impact on IR. In this regard, it is important to recognize that Chat GPT-like large language models (LLMs) are not limited to predicting the next word in a string.
For example, last year a publication in Nature Computational Science reported that an AI trained on the life events of millions of Danes from 2008-2016 was able to predict those who would still be alive in 2020. The accuracy of 80% is well beyond what could have been achieved in any other way. The method, of course, does not simply apply an off-the-shelf LLM; the researchers had to develop a synthetic language for life events. Nevertheless, the study was a breakthrough demonstration of AI in the social sciences. International relations is one of my strongest current interests, and I expect that the next several years will see AI playing an increasingly important role.
I’m especially eager to help develop an interdisciplinary fellows program that will put BU front and center in this arena. The data science center is a unique and ideal environment in which to launch such an initiative.