Collaborative Research: Designing and Assessing Effective "Hands-On" Training for Computational Science

William Marsh Rice University, Houston TX

Investigators

Abstract

Biomolecular simulation is rapidly growing in popularity and scope of application. It is no longer the preserve of a few specialist groups but is widespread - part of the "toolkit" used by researchers in a wide variety of fields, often closely integrated with other experimental research techniques. As the use of biomolecular simulation grows, a corresponding boom in software development is taking place. Giving researchers the right tools and training to develop such software will reduce the extent of 'wheel-reinvention', consequently improve research productivity, and make research more cost-effective. An important question that arises however, is how to we judge the effectiveness of the tools and training used. In order to develop answers to such questions the CECAM Extended Software Development workshop on biomolecular simulations, will be held in Juelich, Germany on October 12-25 2015. This workshop will provide hands-on tutorials incorporating active learning approaches. The workshop will advance the way interdisciplinary computational science training is conducted as well as how tools to support computational science are developed. This grant is being jointly funded by the ACI Division in CISE and by the CHE Division in MPS and will provide travel support for up to twelve US-based scientists to attend this workshop and contribute to achieving workshop objectives. The impact of this workshop will be significant in three directions: It will (i) provide support for US junior and early career scientists to learn a plethora of new tools and techniques central to their science, (ii) lead to an improvement to the software tools being developed as well as the methodology by which the effectiveness of tools is assessed, and (iii) move towards a method of data-driven and dynamic organization of workshops. This provides a natural evolution in the sophistication of simple "hackathons". It will provide a template for how future multidisciplinary hands-on workshops training can be organized. Lessons learned from this workshop will be used as a basis to improve future computational science classes at Rutgers, Rice and elsewhere. The workshop participants will be selected in order to have a broad representation of the communities we want to engage. The PIs will have a special outreach activity to early career scientists (newly established faculty and post-doctoral researchers) as well as graduate students, and will encourage the participation of minorities and women. This workshop will (i) improve the ability of non-computing specialist scientists to utilize existing tools and software better, but also help formulate their needs and requirements better, (ii) enable tool-smiths to answer a critical and recurring question of how to separate the impact and suitability of tools from the effectiveness of the techniques used to teach and train, (iii) develop exercises that will not only improve scientific productivity but also enable domain scientists to understand how to employ tools that reach adequate scales seamlessly from their local environments to remote resources, (iv) propose an agenda and a set of exercises that will be adaptive based upon real-time feedback and assessment, and serve as a template that can be generalized and instantiated to other domains.