Cultivation of Enhanced Bioinformatic-Specific Pedagogical Manipulatives, Interventions, and Professional Development

The education of bioinformatics, as an interdisciplinary science, can be negatively impacted by 1) incoming students attempting to negotiate complex concepts while lacking the educational scaffolding of the field’s constituent disciplines while 2) taking breadth-first course requirements from 3) educators who may be lacking formal pedagogical training. Therefore, we attempted to ameliorate these issues through three different studies. The first study was to identify common areas of code switching observed in required bioinformatics courses and develop software-based manipulatives to minimize the barriers to learning for students. The first manipulative we developed was BAMnostic, its purpose being to make genomic sequencing data used in bioinformatics courses and research more accessible by making the program easily installable across all major operating systems with no external dependencies. The second manipulative we designed was seqlogo. The purpose of seqlogo was to abstract complex software dependencies away from the students and allow them to focus on understanding and exploring sequence motif identification and analysis. As of this writing, both BAMnostic and seqlogo are required software for Department of Computational Medicine and Bioinformatics (DCMB) students at the University of Michigan and have been downloaded 210k and 20k times, respectively. The purpose of the second study was to investigate bioinformatics curricula efficacy by first identifying potential bioinformatic-specific threshold concepts (if any), then suggesting curricular interventions and introducing pedagogical methodologies to address them. Threshold concepts (TC) are defined as troublesome knowledge that is transformative, irreversible, and domain-specific. Through a student-centered approach, the study began with student focus groups and surveys of students affiliated with DCMB to identify problematic concepts within the bioinformatic curriculum. These potential threshold concepts were then refined by direct collaboration with bioinformatics faculty. We received survey responses from 70 bioinformatics students (40% response rate). Students identified five conceptual bioinformatic obstacles: sequential data analysis; statistical distribution(s) identification and application; data ingest, exploration, and management; data scaling; and references to extant knowledge. We collected 19 DCMB faculty survey responses (53% response rate) that suggested that while each identified concept was transformative, troublesome, and fundamental to understanding bioinformatics, none were bioinformatics specific. These findings corroborate other TC interdisciplinary science research suggesting that interdisciplinary fields may not have unique TCs. The final portion of the research was focused solely on designing and developing the Pedagogy of Interdisciplinary Science Education (POISE) training program. The purpose of POISE was to address the gap in professional development specifically regarding the instruction of graduate-level students within the biomedical interdisciplinary sciences. This was a long-term approach to shift professional biases within the biomedical sciences towards a community of practice that supports and incentivizes pedagogical professional development in accordance with cultural evolution theory so that students who expect and respect well-trained educators at the graduate-level will themselves become well-trained educators to future students. This premise serves as a positive feedback loop that could potentially shift academic cultural norms and values. The POISE pilot cohort was comprised of 11 trainees and completed with a 100% completion rate. Trainees identified the most meaningful training as application of learning theories and authentic assessment and evaluation with virtual classroom management and technology in the classroom being the least. These findings and observations are currently planned to become embedded into the Medical Educators Novel Teaching On-Demand Resource (MENTOR) initiative in the University of Michigan Medical School.PhDBioinformaticsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/174340/1/mdsherm_1.pd