Foundations of Biomedical Data Science Seminar Series: Publishing Reproducible and Credible Models

Publishing repeatable and reproducible computational models is a crucial aspect of the scientific method in computational biology and one that is often forgotten in the rush to publish. The pressures of academic life and the lack of any reward system at institutions, granting agencies, and journals means that publishing reproducible science is often either non-existent or, at best, presented in the form of an incomplete description. In this presentation, Herbert will talk about both formal and informal techniques one can use to help ensure that your published work is repeatable. Secondly, he will also discuss what they mean by model credibility and how might one improve the credibility of models, especially those models that are of national importance.  

Herbert Sauro is a Professor in Bioengineering at the University of Washington, Seattle and serves as director of the NIH center on biomedical model reproducibility. Originally educated in the UK, he wrote one of the first biochemical simulators for the PC in the 1980s and made significant contributions to the early development of metabolic control analysis. In the 1990s he started two successful companies to develop simulation software for education and to provide consultancy to the finance industry in the UK. He has received awards for excellence in teaching and a special commendation from the DARPA Information Processing Technology Office. Herbert’s current interests include encouraging best practices in biomedical modeling and investigating new ways to create models that are dependable enough to be used in clinical settings.