SMB MathEd - July 25, 2019 (v2).pdf (3.5 MB)
Cloud-hosted mathematical models: links between education, research, and outreach
Version 2 2019-07-27, 04:12
Version 1 2019-07-25, 15:39
presentation
posted on 2019-07-27, 04:12 authored by Paul MacklinPaul MacklinTo address the challenges facing undergraduate research, we tested lab structure that
can scalably involve undergraduates in cutting-edge mathematical biology research. Teams
of two-to-four undergraduates are jointly mentored by a graduate student (or senior personnel) and the lab’s principal investigator. Graduate students gain extra help and team
management experience, while undergraduates gain modeling experience and co-author scientific papers. Each team is responsible for a new methodology or a new mathematical
model of a specific biological problem. All teams “cross-pollinate” at weekly lab meetings
that combine short progress reports, discussion, and unstructured mentoring time.
We present a case study: a team developed xml2jupyter [1] to automatically convert
command-line agent-based models (written in PhysiCell [2]) to cloud-hosted, interactive
models on nanoHUB [3]. This tool reduced the time to create and deploy a graphical
user interface from months to hours, allowing us to expand the scope and productivity of grant funded projects. Rapidly-developed, cloud-hosted mathematical models have
enabled new approaches to education (adaptive lesson plans; student portfolio pieces),
research (research papers with online model demos), and outreach (sharing interactive
models on social media). With further refinement, we expect that heavy undergraduate research participation will continue to drive unexpected results that benefit education and research. For an example of a cloud-hosted mathematical model [4], see
http://nanohub.org/tools/pc4cancerimmune.