Reproducibility efforts as a teaching tool: A pilot study.

Area of Science:

• Bioinformatics
• Computational Biology
• Scientific Education

Background:

• The scientific method relies on the reproducibility of empirical results.
• A "replication crisis" exists where many research findings are difficult to replicate, threatening scientific credibility.
• Direct replication studies are costly, time-consuming, and often unrewarded in academic publishing.

Purpose of the Study:

• To assess the feasibility of using graduate-level bioinformatics and computational biology courses for replication studies.
• To expose students to real-world biological data analysis challenges, including data acquisition and methodological understanding.
• To evaluate the sufficiency of information in published articles for figure replication by graduate students.

Main Methods:

• A pilot study was conducted in a graduate "Intro to R Programming" course.
• Students were assigned recent scientific articles to replicate figures using original data.
• The course focused on real-world data analysis, including data locating, annotation interpretation, and method understanding.

Main Results:

• Students successfully reproduced 39% of the figures from the assigned articles.
• Key obstacles included insufficient programming skills and incomplete documentation of analytical methods.
• Students demonstrated engagement and enthusiasm throughout the semester.

Conclusions:

• Integrating replication studies into bioinformatics curricula can provide valuable training in the scientific process.
• This approach can enhance understanding of reporting standards and the importance of data openness.
• Supervised replication tasks can potentially lead to fundamental scientific contributions from students.