2025 ASEE Annual Conference & Exposition

The biotechnology and pharmaceutical industries are increasingly reliant on a workforce pipeline of graduates possessing the skills needed to quantitatively describe complex systems to predict functional outcomes relevant to healthy physiological function and to disease states. These skills will be essential for not only identifying novel drug targets and ascertaining the etiology of complex diseases such as cancer and heart disease, but also for achieving truly personalized medical diagnostics, therapies, and surgical approaches toward treating these diseases. Moreover, inherent biological complexity and high-throughput measurement approaches lead to massive “big data” sets, often with thousands of heterogeneous values. This complexity requires data science tools such as data-driven modeling and machine learning to appropriately integrate heterogeneous data. Thus, it is imperative to train a diverse new generation of scientists in the concepts and practice of multi-scale systems bioengineering and biomedical data sciences (BDS) research.

At the University of Virginia, we developed an NSF-funded REU Site in Multi-Scale Systems Bioengineering and BDS (NSF #1560282 & #1950374) that has supported 81 students engaging in research projects for the past eight years (2017-2024). These students were recruited out of a total of 1,375 applicants, with participants drawn from 54 colleges and universities. Two summers (2020 & 2021) the program was run as a virtual REU due to institutional constraints on visiting researchers due to the pandemic. The REU students were matched to a mentor based on a combination of student interest in specific sub-areas of systems bioengineering research and mentor availability each summer. Most research projects relied primarily on previously developed methods and tools and typically involved application to biological data and generation of testable hypotheses. The specific research projects included a wide variety of topics in the field, ranging anywhere from molecular scale biophysics models to cell-scale signaling models, biomedical data science analysis of genetic data, tissue-level biomechanics models, and image analysis algorithms for quantifying cell distribution in tissue-engineered constructs. The participants took part in an introductory bootcamp on the fundamentals of systems modeling and had multiple opportunities to present their research progress throughout the summer to experts in the field. They also received professional development training on research ethics, technical communication, and launching careers in systems bioengineering. We analyzed participant demographics, outcomes in presenting or publishing their work, career outcomes, and survey data from each summer’s cohort.

The 81 REU participants came from 54 colleges and universities and represented 24 different majors, with 47% of them biomedical engineering (BME) majors; 68% were from groups traditionally underrepresented in STEM, 32% were first-generation students; 58% were women; and 41% attended non-R1 institutions; 67% presented their work at national meetings, and nine have become co-authors on ten papers. Of the 60 who have since graduated, 85% are either in graduate school or in STEM industry positions. Post-REU surveys of participants revealed that 98% of respondents rated their overall experience with the REU as either “very satisfied” or “satisfied” (average 4.72 on a Likert scale). Evaluations of specific program objectives and mentoring were similarly high. Regarding impact on long-term goals, nearly 75% said that the REU increased their interest in STEM and encouraged them to pursue further education towards a research or academic career, while 45% said the program helped solidify interest specifically in systems bioengineering.

From a programmatic standpoint, we have several recommendations: Our large number of applications suggest that the specific research area is important, and if well-presented to possible applicants can be a highly motivating selling point. Our experience during the pandemic was that a virtual REU can lead to positive research outcomes, although cohort bonding and the experience of working in a lab are diminished. One challenge of a one-size-fits-all bootcamp is appropriately accommodating the needs of the varied research topics. Some projects require coding, but others do not since they use established software tools; some require model development, others data science. We are continually iterating to find the optimal balance of instruction in topics that support every student in the program.