Report on Software Development Session at 2018 QBI

Written on Mon, Feb 5, 2018 10:00 PM by admin

Discussions about software design, implementation etc. have been an aspect of QBI conferences for many years, with plenary discussions pointing out the importance of software development for the success of the communities activities and the difficulties that many research groups face with software development.


This year's QBI conference was to first to host a minisymposium on software development which consisted of four presentations and was organized and introduced by Raimund Ober. Winfried Wiegraebe from the Allen Institute for Cell Biology presented the variety of software development challenges that their initiative presents due to large amounts of data that need the analyzed. He also pointed out the difference in culture between software developers who have rigorous software development training as it is required in an industrial context and the development efforts as they are often followed in academic research laboratories. Raimund Ober (Texas A&M University) presented their approach to the design of software pipelines and algorithms for single molecule data analysis. Jens Rittscher (Oxford University) discussed the design of a web based portal to use in digital pathology. Mark Tsuchida  (Open Imaging) gave a presentation on issues arising in hardware control in their micro-manager platform.


The session led to various lively discussions related to the role of software development in our community and the fact that software development is handled differently from instrument, algorithm development or most other development activities. In contrast to those other activities, software design and development approaches are rarely discussed at conferences, and the underlying approaches  often remain unpublished. Typically, at the most the result of such efforts, the final product, is presented from a user perspective. This means that the approaches that lead to improved results are hardly ever discussed and little exchange exists between developers, although the problems are often formidable. As a consequence, the role of programmers in research laboratories and their career paths is often different from those pursuing methods development in other fields in a laboratory. The activities of the QBI Society aim to address these issues.