URSSI - SGCI - PresQT: Research Software and Science Gateways: Addressing Sustainability, Usability and Reproducibility Challenges to Enhance Research �
•
1 like•288 views
Research software plays a major role in academia evident in the fast developing landscape of simulations and modeling and new evolving areas such as deep learning. Research software has been gaining increased attention in the last 5 years than ever before. Science gateways are a subgroup of research software addressing the needs for seamless end-to-end solutions so that scientists can focus on their research questions instead of becoming acquainted with details of the often complex underlying computing and data infrastructure. Providers of distributed computing infrastructures such as XSEDE report since 2013 that more users apply their resources for research via science gateways than via command line. Recognizing the importance of software and science gateways, the conceptualization of the US Research Software Sustainability Institute (URSSI) and the Science Gateways Community Institute (SGCI) that is funded since 2016, aim at addressing challenges to achieve sustainability of software and increase the usability of complex computational methods to accelerate research. A cornerstone of science is reproducibility and even though solutions such as science gateways support reproducibility of research, the current landscape needs further tools and concepts to address the reproducibility challenges. One of the goals of the project PresQT (Preservation Quality Tool) is to step into the gap to connect existing systems to achieve easier curation and preservation of data and software and thus, contribute essentially to reproducibility of research. The talk will go into detail for the research software and science gateways landscape in the US, challenges faced by projects and the actions and vision of URSSI, SGCI and PresQT.
Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Similar to URSSI - SGCI - PresQT: Research Software and Science Gateways: Addressing Sustainability, Usability and Reproducibility Challenges to Enhance Research �
Similar to URSSI - SGCI - PresQT: Research Software and Science Gateways: Addressing Sustainability, Usability and Reproducibility Challenges to Enhance Research � (20)
More from Sandra Gesing
More from Sandra Gesing (20)
Recently uploaded
Recently uploaded (20)
URSSI - SGCI - PresQT: Research Software and Science Gateways: Addressing Sustainability, Usability and Reproducibility Challenges to Enhance Research �
- 4. So+ware Sustainability Sustainable so+ware is so+ware which is: • Easy to evolve and maintain • Fulfils its intent over Hme • Survives uncertainty • Supports relevant concerns (PoliHcal, Economic, Social, Technical, Legal, Environmental) (Patricia Lago at WSSSPE4)
- 7. Sustainability for Cyberinfrastructure - NSF Sustainability InsHtutes and Excellence Hubs are funded to support the CI and research community ConceptualizaHons • US Research So+ware Sustainability InsHtute (URSSI) • GeospaHal • … ImplementaHons • Science Gateways Community InsHtute (SGCI) • The Molecular Sciences So+ware InsHtute (MolSSI) • InsHtute for Research and InnovaHon in So+ware for High Energy Physics (IRIS-HEP)
- 12. Areas of Concern • FuncHoning of the individual and team • FuncHoning of the research so+ware • FuncHoning of the research field itself
- 13. FuncHoning of the Individual and Team • Training & educaHon • Ensuring appropriate credit for so+ware development • Enabling publicaHon pathways for research so+ware • Fostering saHsfactory and rewarding career paths for people who develop and maintain so+ware • Increasing the parHcipaHon of underrepresented groups in so+ware engineering
- 14. FuncHoning of Research So+ware • SupporHng sustainability of the so+ware • Growing community, evolving governance, and developing relaHonships between organizaHons, both academic and industrial • Fostering both tesHng and reproducibility • SupporHng new models and developments (e.g., agile web frameworks, So+ware-as-a-Service) • SupporHng contribuHons of transient contributors (e.g., students)
- 15. FuncHoning of the Research Field Itself • Growing communiHes around research so+ware and disparate user requirements • Cataloging extant and necessary so+ware • DisseminaHng new developments • Training researchers in the usage of so+ware • Understanding and improving pipelines of diverse developers and maintainers
- 18. Partner with URSSI We don’t want to reinvent the wheel but partner with exisHng iniHaHves! • UK SSI • So+ware and data carpentries • ACI-REF VR • …
- 19. IniHal Straw Man
- 20. ConceptualizaHon • Workshops • First workshop took place in April in Berkeley • Second workshop took place in October in Chicago • So+ware credit workshop will take place in January in Santa Barbara • Incubator workshop will take place in February in Maryland • Survey with about 1200 answers – in analysis • Ethnographic studies • Mission and vision working group
- 21. How to Engage with URSSI • Watch the website h_p://urssi.us/ • Repos for website and workshops h_ps://github.com/si2-urssi • Blog posts h_p://urssi.us/blog/ • Join the mailing list h_p://urssi.us/ • Discuss h_ps://discuss.urssi.us/ • Twi_er h_ps://twi_er.com/si2urssi • If you have quesHons, want to suggest something, want to volunteer, email us: contact@urssi.us
- 22. Technology-Enhanced Research 22 • Increased complexity of • today’s research quesHons • hardware and so+ware • skills required • Greater need for openness and reproducibility • Science increasingly driving policy quesHons • Opportunity to integrate research with teaching • Be_er workforce preparaHon We need end-to-end solu,ons that provide broad access to advanced resources and allow all to tackle today’s challenging science ques,ons è Science Gateways
- 30. Science Gateway Survey 2014 30 What services would be helpful? • sent out to 29,000 persons • 4,957 responses from across domains • 52% from life, physical or mathemaHcal sciences • 32% from computer and informaHon sciences or engineering • 45% develop data collecHons • 44% develop data analysis tools Proposed Service % Interest Evaluation, impact analysis, website analytics 72% Adapting technologies 67% Web/visual/graphic design 67% Choosing technologies 66% Usability Services 66% Visualization 65% Developing open-source software 64% Support for education 64% Community engagement mechanisms 62% Keeping your project running 62% Legal perspectives 61% Managing data 60% Computational resources 59% Mobile technology 59% Database structure, optimization, and query expertise 59% Data mining and analysis 58% Cybersecurity consultation 57% Website construction 57% Software engineering process consultation 53% Source code review and/or audit 51% High-bandwidth networks 45% Scientific instruments or data streams 44% Management aspects of a project 38%
- 31. Science Gateway Survey 2014 31 34% 36% 20% 17% 31% 26% 42% 16% 30% 18% 45% 44% 14% 15% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Usability Consultant Graphic Designer Community Liaison/ Evangelist Project Manager Professional Software Developer Security Expert Quality Assurance and Testing Expert Wished we had this Yes, we had this Well-designed gateways require a variety of exper-se
- 33. Technologies • Widely used complete frameworks (Galaxy, HubZero, Globus Online etc.) • RESTful APIs and support of mulHple programming languages in widely used frameworks (Apache Airavata, the Agave plazorm, etc.) • Reused interface implementaHons such as the one of CIPRES with its RESTful API (CIPRES has served more than 20,000 users to date) • Science gateways as a service with provision of hardware in the background such as SciGap (Science Gateway Plazorm as a Service) Lessons learned: approaches should be technology agnos-c, using APIs and standard web technologies OR deliver a complete solu-on Community Engagement is key Hubzero instances world wide
- 35. Sustainability via On-Campus Teams On-campus teams It is a centralized team at your insHtuHon – irrespecHve whether you are part of a university, a naHonal lab, an organizaHon, a consorHum or a company… Local teams vs. distributed and remote teams: For local teams it is s-ll easier to build more trust, to be more efficient and to create a strong culture. h_ps://www.codementor.io/blog/modern-engineering-teams-4ea9dp+zs Even ants wish they had an extra pair of hands when developing science gateways! • • • • • • • • • Is your campus seeing an increasing number of research projects that include web-based applications? Does each group have to hire developers independently? This can be time consuming and inefficient. You are not alone. Contact us to request a free consultation, webi- nar, or on-campus visit to start your path toward sustainable gateway development. Addressing Software Sustainability on Your Campus THEREISASOLUTION Synergy between projects http://sciencegateways.org/campusgroups help@sciencegateways.org We can provide supplemental expertise where you don’t have it. We can provide support for your journey to creating a campus-based group. We can provide ongoing advice based on campuses who have successfully created their own groups. Creating a central pool of expertise on your campus offers many benefits including: NOWISTHERIGHTTIME! INTERESTED?CONTACTUS! Science gateways are online, end-to-end solutions that provide broad access to advanced resources. They provide a community space for science and engineering research and education, allowing all to tackle today's challenging science questions. Gateways are an increasingly common component of funded activities by many agencies. Individual PIs find it challenging to recruit and sustain teams that offer the diversity of expertise necessary for developing gateways. HOWTOSTART?
- 36. Connect with SGCI Incubator Sustainability Bootcamp • h_ps://sciencegateways.org/engage/bootcamp • 5 full days • Teams on projects • InteracHvity • Community formaHon • Puwng away the normal daily rouHne • Homework • twice per year • addiHonal ones can be booked (travel expenses for presenters) • adapted to feedback
- 37. Connect with SGCI Incubator Sustainability Bootcamp • h_ps://sciencegateways.org/engage/bootcamp Work with us • h_ps://sciencegateways.org/consulHng/work-with-us Yearly Conference • h_ps://sciencegateways.org/engage/annual-conference Become involved as a partner or affiliate • h_ps://sciencegateways.org/about/partners So+ware/Gateway Catalog • h_ps://catalog.sciencegateways.org/ Train students in internships • h_ps://sciencegateways.org/engage/student-focused Webinars, blogs, newsle_er, Twi_er, LinkedIn etc. h_ps://sciencegateways.org
- 39. Bridging the Gap to Data Sharing Researchers “the local academic community struggles to effectively manage its assets which manifested itself in a number of challenges, and as for researchers, they lacked storage capacity and data curation processes, and the institution lacked standard metadata and indexing technologies, as well as tools that would support the whole research workflow” - Digital Asset Strategy Committee, DigitalND, 2011 Libraries Typically, data curation happens retroactively, and as a result data is either not captured at all or available metadata is incomplete. Pressures from the Outside “...digitally formatted scientific data resulting from unclassified research supported wholly or in part should be stored and publicly accessible to search, retrieve, and analyze.” - White House OSTP Public Access Memo, Feb. 2013 Image Credit Peter Alfred Hess (CC BY 2.0)
- 40. Current Lifecycle of Research Projects Hesburgh Libraries SelecHon/ development of tools Data assembling/ creaHng Reports PreservaHon of Data Funding ends New project Work-intensive and too late in the lifecycle
- 41. Target Lifecycle of Research Projects Hesburgh Libraries SelecHon/ development of tools Data assembling/ creaHng Reports PreservaHon of Data Funding ends New project EASY STEP!!! (ideally) Assure quality of data Assure quality of data
- 42. PresQT A collaboraHve design effort to enhance reproducibility and more open sharing of research data through open source development (July 2018-June 2020) of Tools and RESTful Services to Improve Preserva-on and Re-use of Research Data & SoXware. Hesburgh Libraries h_ps://www.imls.gov/grants/awarded/lg-72-16-0122-16 h_ps://www.imls.gov/grants/awarded/lg-70-18-0082-18
- 43. Hesburgh Libraries h_p://presqt.crc.nd.edu/ Two Workshops & the Needs Assessment answered by 1740 stakeholders have been completed. All Resources avail online
- 45. h_ps://osf.io/d3jx7/ h_ps://cos.io/ Project Partner An open project with all stakeholder input, workshop materials, and meeting info shared on Open Science Framework. PresQT OSF Project
- 46. Needs Assessment Results – over 1700 answers https://ndlib.github.io/PresQTNeeds/
- 48. Repository and Tool Agnostic Solutions • Open design of tools and services using standards • Integrate with workflows, tools, and virtual environments • Priority Focus Areas ➔ Available for anyone to adopt what they need and build upon it!
- 49. Open Design Document • Open design of tools and services using standards • Integrate with workflows, tools, and virtual environments • Priority Focus Areas ➔ Available for anyone to adopt what they need and build upon it!
- 50. Partners and Committed Collaborations • Sheridan Libraries, John Hopkins University • NDS • UC San Diego Library • HUBzero team, Purdue University • Yale University Library • Libraries at Amherst College, Fontbonne University, Tuskegee University, ConfederaHon of Open Access Repositories (COAR) • ReproZip, Jupyter, CERN, RDA groups • Midwest Big Data Hub, Science Gateways Community InsHtute, URSSI, Center for Open Science, Data CuraHon Network, So+ware PreservaHon Network
- 51. Partners and Committed Collaborations • Sheridan Libraries, John Hopkins University • NDS • UC San Diego Library • HUBzero team, Purdue University • Yale University Library • Libraries at Amherst College, Fontbonne University, Tuskegee University, ConfederaHon of Open Access Repositories (COAR) • ReproZip, Jupyter, CERN, RDA groups • Midwest Big Data Hub, Science Gateways Community InsHtute, URSSI, Center for Open Science, Data CuraHon Network, So+ware PreservaHon Network JOIN US!
- 52. Contact us: presqt-contact-list@nd.edu PresQT on the web: https://presqt.crc.nd.edu/ Subscribe to our newsletter! Hesburgh Libraries