Fig 1.
Project summaries and distribution of analysis among students.
Two iterations of the online genomics CURE were taught in the same format and used to conduct 2 completely different research projects. The first 2 modules were dedicated to exploration of the data set being analyzed and featured an introduction to the coding language being used in the course. The middle modules were used to divide the required analysis between the students of the class. For iteration 1, students studied how parameters for quality trimming of RNA sequencing data affected identification of genes differentially expressed by sex in the human placenta. For iteration 2, students used expression of specific sex chromosome genes to infer the sex chromosome complement in cell lines used as models for human cancer. In the last modules, each student described the results of the study in a manuscript which was peer reviewed by other students.
Fig 2.
Instructor and student contributions over 7-week CURE.
Instructor contributions are labeled as circles, student contributions in triangles. Students began by filling out a pre-assessment to show their baseline level of knowledge on the research topic. Each week, students completed research goals and turned them in as assignments along with a progress report communicating their achievements and challenges. Students were assigned weekly scientific writing prompts and publications to read in a journal club designed to help them slowly build up their final reports in Module 6. Instructors used all of these submissions to select topics for weekly lab meetings conducted using teleconference software and recorded for students that could not attend. Instructors hosted shared research (office) hours throughout the week. Instructors monitored progress on research goals and distributed data generated by individual students at the midpoint of the CURE. Following the completion of the course, students completed a learning assessment which instructors analyzed to see which areas students were able to increase their knowledge in as well as misconceptions and struggles that could be avoided in future CUREs.
Fig 3.
Increase in student knowledge and skills after genomics CURE.
(A, B) Boxplots depicting mean student assessment scores before (green) and after (orange) completing the genomics CURE. Each point represents a student who completed both the pre- and post-assessments and the lines connect pre-assessment and post-assessment scores for each student. The mean class score significantly increased from 65.96% to 77.61% (paired t test p-value 0.003, Cohen’s D-statistic 0.79, medium-large effect) for iteration 1 (A) and 61.62% to 65.16% (paired t test p-value 0.00002, Cohen’s D-statistic 0.82, large effect) for iteration 2 (B). (C) Boxplots depicting each pre-assessment (green) and post-assessment (orange) score for all questions divided by topic for CURE Iteration 2: Biology/Statistics, Coding, and Professional Development.
Fig 4.
Challenges across CURE modules (Iteration 2).
(A–C) Word cloud summary of challenges reported in weekly progress reports in 3 main research phases: Module 1 (A) representing the data exploration and coding introduction phase, Module 3 (B) representing the analysis and interpretation phase, and Module 6 (C) representing the reporting results phase. Word use frequency is shown by size and color (larger and darker shade of red for higher frequency).