Linking Results to Publications Using the NDAR Study - Gretchen Navidi

Advancing Autism Discovery Workshop, Gretchen Navidi.

Transcript

1. 1 Linking Results to Publications Using the NDAR Study Advancing Autism Discovery Workshop April 22, 2013 Gretchen Navidi National Database for Autism Research National Institute of Mental Health
2. 2 Agenda  Overview  Definition - What is the NDAR Study?  Purpose - Why is it important?  Location – Where are they?  Policy  Eligibility – Who can create an NDAR Study?  Permissions – Who can access an NDAR Study?  Schedule – When would an NDAR Study be created?  Requirement – Who has to create an NDAR Study?  Fundamentals  Relationship to Publications and Data  Instructions  NDAR Components
3. 3 NDAR Study: Overview  Definition  Allows a researcher to record basic information about the cohort, measures, study types, analysis and results of a study.  Linked to data contained in NDAR  Linked to a publication  Purpose  allow others to understand the analysis in order to:  replicate results  understand the data analysis methods  Location  NDAR public website under the Data from Papers tab
4. 4 NDAR Study: Challenge  NDAR Perspective  Promote transparency  Make published data easily available for re-use and analysis  Allow for research and analysis opportunities for junior researchers  Promote healthy discussion to improve science and advance discovery
5. 5 NDAR Study: Policy  Eligibility  Any user with an NDAR account can create an NDAR Study  Results – published or not – can be promoted  Note: underlying data must be shared in NDAR/federated repository  Permissions  No policy documents required for permission to create an NDAR Study  Since accessing data contained in NDAR or federated repositories, the owner is still bound by other policy documents:  Linking to own data → Data Submission Agreement  Linking to data collected by others → Data Use Certification
6. 6 NDAR Study: Policy  Schedule  Create the NDAR Study when the publication is ‘In Press’  NDAR staff can pull out the information needed from the paper  Share the NDAR Study at the time of publication  Requirement  Researchers are not currently required to create an NDAR Study  Highly encouraged  States  Private  Ability to grant access to others  Shared
7. 7 NDAR Study: Relationship to Publications NDAR Study #2 Contains a subset of data from the Collection used in an analysis Published Paper #2 Based on data in an NDAR Study NDAR Study #1 Contains a subset of data from the Collection used in an analysis Published Paper #1 Based on data in an NDAR Study NDAR Collection Contains all data for a grant • Phenotypic Data • Neuroimaging • Genomic  If submit data to NDAR:  Create a Collection → Submit all research data  Analyze  Publish  Report results in NDAR Study → Link to publications
8. 8 NDAR Study: Relationship to Publications NDAR Study #1 Contains a subset of data from the Collection used in an analysis Published Paper #1 Based on data in an NDAR Study NDAR Collection 1 • Phenotypic Data • Neuroimaging NDAR Collection 2 • Phenotypic Data • Neuroimaging • Genomic  If Access data in NDAR:  Query > 1 Collections/Repositories 8→ Retrieve Results  Compute/Process → Load and Analyze  Publish  Report results in NDAR Study → Link to publications
9. 9 NDAR Study: Examples
10. 10 NDAR Study: Examples
11. 11 NDAR Study Components  General Information  Cohort  Measures  Study Types  Data Analysis  Results  Permissions
12. 12 NDAR Study: General Information 1
13. 13 NDAR Study: General Information 2
14. 14 NDAR Study: Cohort
15. 15 NDAR Study: Cohort
16. 16 NDAR Study: Cohort  Data can be associated with an NDAR Study by associating subject GUIDs when creating the NDAR Study.  Option to narrow down the measures (data structures) and even the measure related data elements used in the analysis
17. 17 NDAR Study: Measures  Data structures will automatically be populated based on data for GUIDs already associated with the defined cohorts  Select data structures used in analysis
18. 18 NDAR Study: Measures  Select data structures used in analysis  Select specific data elements used in analysis  Define the measure/data structure and data elements as either primary or secondary.
19. 19 NDAR Study: Study Types  Controlled or observational  Options for the study description, study arms/ comparison groups and intervention type are customized depending on whether the type is controlled or observational.
20. 20 NDAR Study: Data Analysis  Further information options become available depending on the type of analysis
21. 21 NDAR Study: Data Analysis  Further information options become available depending on the type of analysis
22. 22 NDAR Study: Results  Upload supporting documentation or provide urls for websites with more information.
23. 23 NDAR Study: Permissions  Provide permissions to other users while NDAR Study is in a private state.  Read Only  Administrator
24. 24 NDAR Study: Register NDAR Study with Publication
25. 25 NDAR Study: Linking Publications and Data
26. 26 Data from Papers
27. 27 Data from Papers Direct access to the NDAR Collection for users with access to the NDAR Portal. List of project investigators. Study abstract Link to study results/publication.
28. 28 Data from Papers Direct access to the NDAR Collection for users with access to the NDAR Portal. List of project investigators. Study abstract Link to study results/publication.
29. 29 Data from Papers Study cohorts defined according to NDAR methods. Data used in the analysis which can be linked back to data structures in the NDAR Data Dictionary Analysis methods defined
30. 30 NDAR Study Features  NDAR Philosophy  Develop foundation and work with community to extend  Example: NDAR data dictionary started with ~20 measures and has been extended over the past 5 years by NDAR and the community  All components have been pre-populated yet allow for new options to be added  Encourage community to notify us about new methods and options for analysis, tools, etc to keep the NDAR Study feature up to date
31. 31 NDAR Study: Planned Features  Ability to share the NDAR Study while still maintaining some or all of the underlying data in a private state.  Ability to query the NDAR Study  Ability to query by Data Element  Registration and linkage of publications using the Digital Object Identifier (DOI)
32. 32 Discussion Points  Grant Related Issues  Should this be an expectation for all published papers?  Should this be added to the terms and conditions of award?  Should this and other data submission activities be part of the grant application review process?  Schedule Issues  Is the schedule for creation of and sharing of an NDAR Study acceptable?  Technical Issues  Areas for Improvement  How can we make it easier to use  Provide a template  Methods to increase awareness and adoption of use
33. 33 Discussion Points  Marketing  Methods to increase awareness and adoption of use  Work with publishers to require NDAR Study and link in publication  Open up NDAR Study to journal reviewers to review a paper and underlying data prior to acceptance  Create a culture shift  Data sharing  Transparency  Collaboration  Replication of results
34. 34 Contact Information  Public Website  http://www.ndar.nih.gov  NDAR Study  http://ndar.nih.gov/data_from_papers.html  Help Desk  NDARHelp@mail.nih.gov  Gretchen Navidi (NDAR)  navidig@mail.nih.gov