This webinar will provide an overview of some efforts underway at the National Center for Biotechnology Information to help improve scientific reproducibility in genomics research.
Scientific reproducibility relies on literature, data, and code. While biological literature is becoming more open, and data is becoming more widely shared, code written in a biomedical context is often not reproducible.
In many scientific disciplines, the rapidly emerging datasets exceed the available resources of traditional software development communities. Fueled by this shortage and aided by increasingly powerful scripting languages, domain experts often turn to their own devices, which can create difficulties in reproducibility because of issues with documentation, ongoing development, and advertising.
This session will describe efforts NCBI has undertaken to address these issues, including NCBI-facilitated hackathon programs (with some teams working on modular, reproducible workflow generators), “analyze-athons,” reproducibility workshops, student discovery challenges, and the scale-up of data science training worldwide.
For more information on other webinar in this series, click here.