+Deep learning for structural biology has numerous applications, ranging from basic research through to clinical application. A key challenge with these approaches is accessibility, and much work is currently being conducted to develop computational workflows to improve this access. However, building, executing, and sharing workflows is challenging. The difficulty only increases when implementing novel approaches like deep learning, and even more so when moving these workflows to national infrastructure to operate at scale. Some of the challenges include knowing how to develop a workflow to optimise use of these large compute systems, how to configure and deploy workflows that can efficiently employ both CPU and GPU, and how to share the workflows created so that others can find, understand, reuse, and ultimately cite the workflow. This workshop will examine these challenges from the perspective of a workflow’s life cycle, and in doing so provide both theory and hands-on activities that will support bioinformaticians to develop, execute, and share computational workflows for structural biology at scale.
0 commit comments