Presentation is loading. Please wait.

Presentation is loading. Please wait.

A cloud platform for interactive reproducible computational experiments Siddeswara Guru s.guru@uq.edu.au Data Science Director.

Similar presentations


Presentation on theme: "A cloud platform for interactive reproducible computational experiments Siddeswara Guru s.guru@uq.edu.au Data Science Director."— Presentation transcript:

1 A cloud platform for interactive reproducible computational experiments
Siddeswara Guru Data Science Director

2 TERN1 Purpose National infrastructure for collecting, storing and sharing Australia’s terrestrial ecosystem datasets and knowledge. Build and manage data infrastructure to provide public access to terrestrial ecosystem data. Facilitate open access to terrestrial ecosystem research data. 1TERN is supported by the Australian Government through National Collaborative Research Infrastructure Strategy (NCRIS)

3 Motivation reusable data  reusable science
Ease of access to compute platform

4 Reproducible Science Duplicate the scientific experiments or reproduce experiment results. Our focus is on computational reproducibility. Source: nature.com

5

6 Challenge Lack of culture to make scientific claims reproducible
Lack of detailed information about data and code even after they are published “lack of an integrated infrastructure for distributing reproducible research to others”1 In US, $28 billion per year spent on clinical research that are not reproducible2. Source: experimentalmath.info 1Peng, Roger D. (2011). Reproducible Research in Computational Science. Science, 334(6060), doi: /science 2Freedman, L. P., Cockburn, I. M., & Simcoe, T. S. (2015). The Economics of Reproducibility in Preclinical Research. PLoS Biol, 13(6), e doi: /journal.pbio

7 Reproducibiltiy develop an infrastructure where computational experiment e that has been developed at time t on a hardware and software infrastructure h using data d is reproducible at time t1 on same hardware and software infrastructure h using the same data d.

8 Integrated Infrastructure
"one of the most effective ways to promote high-quality science is to create free open-source tools that give scientists easier and cheaper ways to incorporate transparency into their daily workflow:"3 Cloud-based platform Scientific workflows executable on a easily accessible platform 3 Stuart Buck, Solving reproducibility, Science 26 June 2015: 348 (6242), [DOI: /science.aac8041]

9 System Developed Cloud-based virtual desktop with underlying storage, applications and data over a web browser.

10 Case Study: IUCN Red List Ecosystem Assessment

11 International Union for Conservation of Nature (IUCN) Red List of Ecosystem Framework
Rodrı´guez JP et al A practical guide to the application of the IUCN Red List of Ecosystems criteria. Phil. Trans. R. Soc. B 370 :

12 Workflow

13 Criteria A Several more examples are built

14 Moving forward Currently, all artifacts are encapsulated in the infrastructure. Each of the artifacts could be described in linked data principle. Workflow, a research based asset invoked as WfaaS.

15 Initial Project Sponsors
collaborators Initial Project Sponsors Acknowledgements : Hoang Nguyen, Yi Sun, Ivan Hanigan, Emma Burns, Tim Clancy and many others The Workflow shown is accessible from Thank you Access CoESRA Desktop Register/login Access Virtual desktop


Download ppt "A cloud platform for interactive reproducible computational experiments Siddeswara Guru s.guru@uq.edu.au Data Science Director."

Similar presentations


Ads by Google