Research Partners to Coordinate Domain-Specific Data Services Medical Library Association Webinar Data Services Specialization - Advanced Level September 30, 2021
BACKGROUND BS Marine Biology, UNH MLIS, Simmons University Library Fellow, UMass Medical eScience Coordinator, NNLM NER Data Services Librarian, Harvard Data Science Certificate, edX DATA SERVICES ▪ Consultations ▪ Teaching and Training ▪ Resource Development ▪ Embedded Services ▪ Carpentries Instructor ▪ Journal Managing Editor
POW HARVARD LIBRARY ▪ Umbrella organization for Harvard’s libraries, archives, and shared services ▪ Highly decentralized ▪ 70+ libraries and archives ▪ Supporting 11 schools & institutes ▪ Research Data Management Program (3 FTE) LONGWOOD MEDICAL AREA ▪ Includes Harvard Medical School, Harvard School of Dental Medicine, Harvard T.H. Chan School of Public Health, and 16 Affiliate Teaching Hospitals ▪ 12,400+ faculty & 10,000+ fellows ▪ Various degree programs ▪ Publishing & Data Services Team (3 FTE, 2 PTE)
1. Are you currently a data librarian or involved in data services? 2. What do you love about working with data or data management? OR What do you want to know about working with data? 3. What is a food that everyone seems to love but you dislike?
▪ Explain the importance of coordinated data services ▪ Identify domain-specific research data practices ▪ Tailor data services to local community needs
and ongoing management of data through its lifecycle of interest and usefulness to scholarship, science, and education.” — The University of Illinois’ Graduate School of Library and Information Science Research Data Management
DATA COLLECTION PRACTICES Organize files Naming conventions Maintain documentation Project version control DATA SHARING PRACTICES Connect research Share in a repository Close out project Research Data Lifecycle by LMA RDMWG
Research Data Lifecycle by LMA RDMWG DATA SHARING PRACTICES Library assists with curating, sharing, and depositing data in a repository Research Administration ensures compliance with regulations & policies DATA COLLECTION PRACTICES Library helps with data organization and documentation practices Research Computing assists with storage of data files and offers classes on Github
Partners in Data Services • University Information Technology (HUIT) • Library Technology Services (LTS) • Faculty of Arts & Sciences Research Computing (FAS-RC) • Institute for Quantitative Social Sciences (IQSS) • Office of the Vice-Provost for Research • Offices of Sponsored Research (2)
Three Levels of Supports: ▪ Effective Service Team ▪ Research support services including the library, IT, office of sponsored programs, etc. ▪ Supportive Organization ▪ Strong service provider network with lead and partner agencies ▪ Hospitable System ▪ Supported by funding and policies Effective Service Team Supportive Organization Hospitable System
Services ▪ Libraries are important providers of research data services ▪ IT/research computing provide fewer research data services than libraries, but are also an important provider ▪ A wide variety of services are provided by academic departments, research centers and facilities, and professional schools ▪ Libraries are leaders in developing institutional data policy and research data services ▪ Institutions offer more mature services when libraries take a lead role rather than a supporting role ▪ Institutions offer more mature services when more on-campus stakeholders are engaged during development Research Data Services in US Higher Education (2020) Determining research data services maturity: The role of library leadership and stakeholder involvement (2021)
Coordinating Services ▪ A recent increase of research data and computing services: ▪ Along with increase in data-centric and data science research ▪ To support funders and journals requirements ▪ But services not funded at same pace ▪ Uncoordinated growth resulting in services distributed across units and schools often disconnected (decentralized and uneven)
Challenges ▪ Storage ▪ Next-Generation Sequencing ▪ Image data ▪ Big & Unstructured data ▪ Documentation ▪ Electronic Lab Notebooks ▪ Standards (file, metadata, etc.) ▪ Sharing ▪ Sensitive Data ▪ Personally Identifying Information ▪ Ethical considerations 1. Adapting data management education to support clinical research projects in an academic medical center (2019) 2. Big data management challenges in health research—a literature review (2019) 3. Anonymising interview data: challenges and compromise in practice (2015) 4. The impact of next-generation sequencing on genomics (2011) 5. Issues in Biomedical Research Data Management and Analysis: Needs and Barriers (2007)
Establish collaboration between Research, Library, and IT/Research Computing ▪ Build a catalog to learn what services are provided across units: ▪ Standardize the information ▪ Unify services when possible ▪ Create a site to find all service offerings in a common way ▪ Find gaps and connect services, tools, and teams ▪ Foster a community of research computing and data teams
Details Staff Development at Harvard Library - Recommendations for Supporting Strategic Initiatives (2018-11) 27 library units surveyed about their current and future data services plans
PLANNING • Data Management Plans (DMPs) consultation, review, training • Referrals to IT & Research Computing • Navigation of institutional RDM services and resources COLLECTION & STORAGE • Electronic lab notebook support • Metadata and data documentation services • File organization and asset management training ANALYSIS • Referrals to IT, Research Computing, and research support services • Promotion of research support services • Comparisons of data cleaning, analysis & visualization tools REUSE • Locate data for new project/data discovery • Best practices for data reuse training • RDM lifecycle training SHARING • Data repository administration • Referrals to specialized subject data repositories • Data sharing and publishing training (e.g. DOIs, data citations) • Data use agreements (DUAs) ARCHIVING • Archival processing & appraisal • Digital preservation and stewardship services • Digital preservation and stewardship
Checklists ▪ Tools used to better organize your projects and to verify your most important tasks ▪ Help you get started integrating data management into your research practice ▪ Identify gaps and communicate elements of data management to your team ▪ Apply to an individual research project ▪ Revisit periodically, adopting new guidance, procedures, and techniques
Prompts 1. What is the Data Services Model at your institution? 2. Have you identified RDM challenges in your domain? 3. Have you identified stakeholder partners to work with? 4. Have you used or developed checklist resources for RDM before?
& Security DSL5 - Sensitive Data that could place the subject at severe risk of harm Storage: requires security consulting for special handling DSL4 - Sensitive Data that could place the subject at significant risk Storage: Harvard Secure Transfer, External hard disk with encryption, RedCAP DSL3 - Sensitive Data and some regulated data that could be damaging Storage: Harvard Dropbox, Shared network, OneDrive, SharePoint DSL2 - Unpublished non-sensitive research data Storage: Harvard standard email DSL1 - Publicly available and unrestricted data Storage: Public repositories, consumer products
Bringing new people into a project ▪ Anticipating data management needs for a data management plan ▪ Changing data management practices and improving training ▪ Understanding when Data Use Agreements are required ▪ Utilizing metadata standards ▪ Storing and retrieving large datasets
Onboarding ▪ Set-up standard operating procedures for data management ▪ Facilitates accurate communication about projects ▪ Checklist outlines the important steps for onboarding new team members, recommended data management practices, and Harvard policies and resources RDM Onboarding Checklist
▪ Seminars ▪ 5 in-person sessions (since April 2019) to 71 attendees ▪ 1 virtual session (September 2020) with 90 attendees ▪ Customization ▪ Genetics Department ▪ “The transition to the new checklist has been seamless, and feedback from lab members on the new forms has been overwhelmingly positive.” – Lab Data Manager
Plans ▪ Document outlining plans for handling all of the data while in the lab and a plan for data to leave the lab (dissemination) ▪ Can be formal or informal depending on circumstance ▪ Living document that is frequently referred to and updated as needed Types of data (source, format, size) Documentation and metadata (e.g., reagent, technical, experimental, analytical, dataset level) Storage, backup, and security Privacy and confidentiality protections Plans to disseminate Access and sharing Preservation or destruction Roles and responsibilities ELEMENTS OF A DMP
Template GUIDANCE & RESOURCES A free tool that helps researchers create data management plans that fulfill the requirements of different funders. 1. Go to https://dmptool.org 2. Login using your Harvard Key 3. Use a Harvard DMP Template a. Complete Instructions
Text “All human-subjects data submitted to the repository will include a Global Unique Identifier (GUID) but will not include personally identifiable information (PII). Submission of data into and sharing of data via the planned repository will be mentioned in the informed consent process of the study, and the awarded institution and Principal Investigator will ensure that submission and sharing via the repository are consistent with the informed consent of study participants from whom the data are obtained. All data collected on human subjects involved in this NIH-supported clinical trial will be submitted, including data from control subjects. All data made available via the repository will be de-identified data, such that the identities of participants cannot be readily ascertained or otherwise associated with the data by the repository staff or secondary data users. Investigators and institutions seeking data from the repository will be expected to meet data security measures and will be asked to submit a data access request, including a Data Use Certification, which is co-signed by the investigator and the designated Institutional Official(s) at the NIH-recognized sponsoring institution with a current Federal Wide Assurance (FWA).”
VIDEO TUTORIAL ▪ Seminars ▪ 7 (since July 2018) to 209 attendees ▪ Consultations ▪ 5 DMP/DMPTool since 2020 ▪ “I want to share with you that the data management plan, which you reviewed for me, has passed! I am so happy! I want to thank you for all your help! Your comments were so extremely helpful, and the DM plan would not have passed without your feedback!” – Data Analyst, Department of Epidemiology
Types of documentation to capture for biomedical data includes: ▪ Reagent: Information about the clinical samples, biological reagents (e.g. cell lines, antibodies, siRNAs), chemical reagents (e.g. drugs), etc. used to generate the data. ▪ Technical: Information automatically generated by research instruments and associated software. ▪ Experimental: Information about the experimental conditions (e.g. assay type, time points), the experimental protocol, and the equipment used to generate the data. ▪ Analytical: Information about data analysis methods including software name and version, quality control parameters, and output file type details. ▪ Dataset Level: Information about the objectives of the research project, participating investigators, relevant publications, and funding sources.
Results STANDARDS EXAMPLES ▪ Small molecule, protein, cell line, and antibody standards developed by LINCS consortium ▪ NIH LINCS, NCI HTAN, Open Microscopy Environment ▪ ISO8601 for date; unique ID across the lab; metadata template for technicians Biomedical Metadata Guides
to Coordinated Services ▪ Onboarding Support: Checklist Resource ▪ Partners: Library, Archives, Research Computing, Research Operations, Academic Research and Integrity ▪ DMP Review Service: Custom Templates ▪ Partners: Partners: Library, Research Computing, Research Administration, Research Operations ▪ Metadata Standards: Metadata Guides & Survey ▪ Partners: Partners: Library, Archives, Research Computing, Research Operations, Academic Research and Integrity
Prompts 1. Do you have coordinated services for RDM planning? 2. If you do, what do they look like? How would you rate the success? 3. If you do not, what area would you like to develop and how?
▪ Difficulty of organizing, storing, and retrieving data ▪ Managing large datasets ▪ Need for improved methods of analyzing and processing data ▪ Maintaining documentation throughout a project ▪ Limited availability of institutionally provided expertise and systems
Documentation ▪ Written at many “levels” and comes in many forms ▪ Record all the information necessary to understand the content and context of the data ▪ Stored alongside your research data such as in lab notebooks, databases, or in README Files Cornell Research Data Management Service Group. Guide to writing "readme" style metadata template
Tools ▪ Open Science Framework: Project management tool to manage files in one centralized location and easily build custom organization for your project ▪ Protocols.io: Platform that allows you to create step-by-step detailed, interactive, and dynamic protocols that can be versioned, verified, and published ▪ Addgene: Plasmid repository that facilitates the exchange of genetic material between laboratories, and curate plasmid collections for specific purposes ▪ GitHub: Hosting for software development and version control using Git ▪ Jupyter Notebook: Computational environment used to combine software code, computational output, and explanatory text (i.e. documentation) ▪ R Markdown Reports: Use R Markdown to turn your analyses into documents, reports, presentations and dashboards
Notebooks @Harvard eLABJOURNAL ▪ For documenting, organizing, and searching research data ▪ Efficiently track any sample, specimen, chemical, or instrument in the lab ▪ HMS On-quad Labs, Departments, and approved HMS Cores RSPACE ▪ Collaborative tool to work together in a centralized, streamlined, and secure way ▪ Enables efficient lab management of labs, projects, and teams ▪ Harvard University Schools & Departments (fee based service) Electronic Lab Notebook Guide
to Coordinated Services ▪ Documentation: READMEs & Metadata Guidance ▪ Partners: Library, Archives, Research Computing, Research Operations, Academic Research and Integrity ▪ Active Research Trainings: Data & Script Management ▪ Partners: Partners: Library, Archives, Research Computing, Postdoc Office, Academic Research and Integrity, Bioinformatics Core ▪ Electronic Lab Notebooks: Metadata Guides & Survey ▪ Partners: Partners: Library, Research Computing, Research Operations, Academic Research and Integrity
Discussion Prompts 1. Do you have coordinated services for active data management? 2. If you do, what do they look like? How would you rate the success? 3. If you do not, what area would you like to develop and how?
Researcher attitudes towards sharing ▪ Determining what data will be shared, to who, when, and where ▪ Using non proprietary approaches ▪ Patient confidentiality and privacy concerns ▪ Lack of infrastructure for records retention periods
Infrastructures DRS ▪ Harvard Library Preservation Services ▪ Managed preservation & access repository ▪ Library-facing; not for public deposit ▪ 10+ million objects ▪ 100+ million files DATAVERSE ▪ Institute for Quantitative Social Sciences (IQSS) ▪ Free data repository Open to worldwide research community ▪ Multi-disciplinary ▪ Access to 108,000+ datasets in DV network DASH ▪ Harvard Library Office for Scholarly Communication ▪ Open Access Institutional repository for Harvard research ▪ 50K+ works ▪ 38M+ downloads
▪ Harvard Dataverse Repository provides fee-based curation services to researchers ▪ Ensure datasets are FAIR: findable, accessible, interoperable, and reusable Curation and Data Management Services
▪ Capture institutional and project knowledge ▪ Plan for project member turnover before it happens ▪ Ensures projects can continue following departures ▪ Enables reproducibility RDM Offboarding Checklist
File ▪ Key for proper Offboarding ▪ Document essential information related to projects and datasets ▪ Facilitates data sharing and and reuse, and increases visibility of research RDM Knowledge Transfer File Template
KTF Services ▪ Seminars ▪ 4 sessions (since September 2019) to 76 attendees ▪ About the KTF: “EVERY LAB NEEDS THIS CONTENT! I really appreciate the level of specificity and examples used for each type of research product.” – Lab Data Manager ▪ Customization ▪ Genetics Department
to Coordinated Services ▪ Data Sharing: Dataverse Data Curation Services ▪ Partners: Library, Institute for Quantitative Social Sciences, Information Technology, Office for Scholarly Communication ▪ Offboarding Support: Checklist Resource & Knowledge Transfer File Template ▪ Partners: Library, Archives, Research Computing, Research Operations, Academic Research and Integrity
Prompts 1. Do you have coordinated services for data dissemination? 2. If you do, what do they look like? How would you rate the success? 3. If you do not, what area would you like to develop and how?
Checklist ▪ The Research Data Management Lifecycle Checklist serves as a reference document to keep track of the elements that make up good research data management in the RDM lifecycle ▪ Use the checklist as a guide to find where you are in the lifecycle and to keep track of good research data management elements Biomedical Research Data Management Checklist
Collaboration is Key Universities need to collaborate between research, libraries, and IT: ▪ Catalog all the services offered to support research data ▪ Centralize costly and complex services ▪ Build workflows to integrate services, research tools, computing, and repositories
1. GET TO KNOW YOUR CAMPUS • Who is your target audience? • What resources are already available? • Develop partnerships 2. FILL IN GAPS IN SERVICES • What value can you add? • Know your limits 3. LEAD BY EXAMPLE • If you do cool things, they will come 4. RUN / PLAN / HOST • People will (sometimes) come • Find partners 5. BE OK WITH REJECTION • “That’s not what libraries do.” 6. KEEP AN OPEN MIND • Unexpected opportunities can be fruitful
Discussion Prompts 1. How can academic libraries coordinate high quality, domain-specific data services? 2. What lingering questions or challenges do you face for coordinating services?
Reading List ▪ Baker, D. 2020. Citizen Scientists Are Helping Researchers Design New Drugs to Combat COVID-19. HHMI News. https://www.hhmi.org/news/citizen-scientists-are-helping-researchers-design-new-drugs-to-combat-covid-19 ▪ Borghi J, Abrams S, Lowenberg D, Simms S, Chodacki J (2018) Support Your Data: A Research Data Management Guide for Researchers. Research Ideas and Outcomes 4: e26439. https://doi.org/10.3897/rio.4.e26439 ▪ Briney KA, Coates H, Goben A (2020) Foundational Practices of Research Data Management. Research Ideas and Outcomes 6: e56508. https://doi.org/10.3897/rio.6.e56508 ▪ Corpas M, Kovalevskaya NV, McMurray A, Nielsen FGG (2018) A FAIR guide for data providers to maximise sharing of human genomic data. PLoS Comput Biol 14(3): e1005873. https://doi.org/10.1371/journal.pcbi.1005873 ▪ McKiernan, EC, Bourne, PE, Brown, CT, Buck, S, Kenall, A, Lin, J, McDougall, D, Nosek, BA, Ram, K, Soderberg, CK and Spies, JR, 2016. Point of view: How open science helps researchers succeed. elife 5: e16800. https://doi.org/10.7554/eLife.16800 ▪ Wilkinson, M, Dumontier, M, Aalbersberg, I, et al. 2016. The FAIR Guiding Principles for scientific data management and stewardship. Sci Data 3: 160018. https://doi.org/10.1038/sdata.2016.18 ▪ Wilson G, Bryan J, Cranston K, Kitzes J, Nederbragt L, Teal TK (2017) Good enough practices in scientific computing. PLoS Comput Biol 13(6): e1005510. https://doi.org/10.1371/journal.pcbi.1005510
goldmj
drpoonamkc
hiragahh
juanmartin2026
pawan_pharm
bunnchinn3
robintux
aokiplayer
2720japanoke
naochim2026
bvr7352
3l4l5
inesmontani
shlominoach
tammyeverts
addyosmani
jonoalderson
427marketing
reverentgeek
scottboms
helmedeiros
aleyda
lauravandoore
tapps
