Mr Aswin Narayanan1,4, Dr Steffen Bollmann2, Mr Mark Endrei3, Mr Joshua Scarsbrook2
1Australian Institute for Bioengineering and Nanotechnology, The University of Queensland, Brisbane, Australia, 2School of Electrical Engineering and Computer Science, The University of Queensland, Brisbane, Australia, 3Research Computing Centre, The University of Queensland, Brisbane, Australia, 4National Imaging Facility, Brisbane, Australia
Biography:
Aswin Narayanan (orcid.org/0000-0002-4473-7886):
Aswin Narayanan is the Imaging Informatics Fellow at the University of Queensland Node of the National Imaging Facility (NIF). Aswin is based at UQ's Centre for Advanced Imaging and works on data management and analysis solutions for biomedical imaging, including modalities such as MR, PET, CT, and radiochemistry. With a background in Biomedical Engineering, Aswin has over 15 years of experience supporting scientific research and building software research infrastructure.
Steffen Bollmann (orcid.org/0000-0002-2909-0906):
After a PhD on multimodal imaging at the University Children’s Hospital and ETH Zurich, Switzerland, Dr Bollmann joined the Centre for Advanced Imaging at the University of Queensland (Australia), where he pioneered the application of deep learning methods for quantitative susceptibility mapping. In 2019 Dr Bollmann joined the Siemens Healthineers collaborations team at the MGH Martinos Center in Boston where he worked on the translation of deep learning reconstruction techniques into clinical applications. Since joining the School of Information Technology and Electrical Engineering at the University of Queensland in 2020 Dr Bollmann develops computational methods to process magnetic resonance imaging data.
Mark Endrei (orcid.org/0000-0001-9368-2773):
Mark Endrei is a senior principal research software engineer at the Research Computing Centre, The University of Queensland, Australia. He also has more than 20 years of experience in IT industry, working with large corporations both nationally and internationally. He has a PhD from The University of Queensland and a Bachelor of Engineering Degree (H1) in Computer Systems Engineering from RMIT University.
Joshua Scarsbrook (orcid.org/0000-0002-0071-5466)
Joshua Scarsbrook is Research Officer with 10 years of professional research experience. Their passion lies in developing solutions to tackle the ever-growing challenges in the field of operating systems.
Throughout their career, they have been involved in numerous projects that have allowed me to hone my skills in the areas of visualization, virtual machines, and operating systems. Their work has been focused on developing innovative approaches to address complex issues across interdisciplinary fields.
Abstract:
Background
The analysis of scientific data necessitates specialised software and complex processing pipelines. Researchers and research infrastructure maintainers expend significant time and effort on software compilation, installation, and managing deployments across diverse computing platforms (laptops, workstations, HPCs, cloud). Additionally, system-specific dependency issues and challenges in publishing reproducible pipelines alongside data often compromise research outcomes.
Introduction
SCIGET aims to deliver a robust solution tackling these challenges in scientific software accessibility and research reproducibility. Leveraging containerisation and projects like Neurodesk.org, CernVM-FS, eessi.io, and tinyrange, SCIGET provides a robust software distribution system alongside accessible and portable virtual desktops that can be co-located with data for scientific analysis.
Methods
At the core is a community-engaged build pipeline, where scientific applications and reference datasets are proposed, packaged, security scanned, and published to container registries. Applications are accessible via graphical virtual desktops embedded into JupyterLab, and command line as modules. Containers run on infrastructure through a variety of mechanisms: Apptainer/Singularity and CernVM-FS on HPCs; Kubernetes on research and commercial cloud; Docker/Podman or unprivileged QEMU VMs for researchers. A comprehensive metadata database provides easy tool discovery. In collaboration with the National Imaging Facility, SCIGET is integrated into the Australian Imaging Service, offering biomedical imaging researchers a comprehensive data and analysis solution.
Conclusion
SCIGET aims to accelerate scientific discovery by streamlining access to essential software. It reduces setup overhead, enhances reproducibility across computing platforms, and fosters greater collaboration by ensuring consistent analytical environments. SCIGET is domain-agnostic and seeks to expand across scientific domains via community collaboration and engagement.