AUG Meeting Jan 31, 2014

6 posts
Was this reply useful? Learn more...
[-] DanMobley +0 points · about 3 years ago

National Sleep Research Resource (NSRR)

Academic User’s Group Meeting and Conference Call Minutes

January 31, 2014 3:00 p.m. EST

NSRR Participants

  • Susan Redline
  • Remo Mueller
  • Michael Rueschman
  • Susan Surovec
  • Daniel Mobley
  • Ary Goldberger

AUG Members

  • Florian Chapotot
  • Nalaka Gooneratne
  • Daniel Gottlieb
  • Craig Johnson
  • Paul Peppard
  • Katie Stone
  • Simon Warby
  • James Wilson
  1. Introduction/Welcome to the AUG Susan R. opened the meeting and provided a brief introduction of NSRR staff members from Brigham and Women’s Hospital, Beth Israel Deaconess Medical Center, and Case Western Reserve University. A brief introduction of the AUG members was then provided.

  2. NSRR Overview Susan R. provided an overview of the NSRR grant main goal, and discussed collaboration with BioLincc and possible future collaboration with dbGaP. Funding for NSRR is via a 5 year NIH/NHLBI grant with support from Brigham and Women’s Hospital (BWH) in years 6 through 10. The NSRR Aims were described (summarized below), and an NSRR Organizational chart was presented with an explanation of the working groups/teams.

    • a. Aim 1—Sleep-Arch: Electronic data library containing raw signals, scored annotation files, sleep summary statistics curated from 15 large-scale research studies of adults and children. Also will include time and frequency domain analyses for ECG and EEG signals from the PSGs, and will link PSGs to covariate data and spectral analysis results.
    • b. Aim 2—Sleep-Terms: Create a standardized infrastructure to support terminology across the data library with mappings of terms and covariate data to a common terminology.
    • c. Aim 3—Sleep-Port: A web-based portal to access, search and visualize the data library using a cloud-based platform. This resource is currently hosted at BWH in Boston but will eventually be cloud-based.
    • d. Aim 4—Sleep Tools: Provide a suite of data curation, data integration and signal processing tools to facilitate offline analysis and discovery of associations among physiological systems and clinical outcomes by investigators. The plan is to provide tools to normalize/standardize EDF files and allow viewing of physiological signals collected using diverse proprietary platforms, and tools to conduct offline spectral analysis as well as advanced processing of ECG, EEG, and breathing signals.
    • e. Aim 5—Support the research community.
  3. Purpose of the AUG Susan R.—Academic User Governing Committee is comprised of key stakeholders who are representatives of major participating study groups, community users, and technical experts. The AUG will meet bi-monthly via conference call and 1 to 2 times annually in person. The AUG will develop governance policies, provide input into usability, accelerate the dissemination of tools, help prioritize tasks, provide feedback on web-based user guides and webinars, and create a dynamic user community.

  4. Unveiling of Remo M. created a survey requesting feedback from AUG members after the meeting and explained the 3 components of Explore datasets, download data, analyze data.

    • a. site demo: Remo provided a demo of the site that included review of uploaded SHHS documentation, variable searching and site navigation, downloading EDF data as single files or full set of files using WGet, demo of annotation files and description of how to visualize annotations, review of posted forms used during study collection, and an overview of tutorials.
    • b. Tools and Resources: Remo gave an overview of the “Tools” page and used the Block EDF Loader tool as an example of a tool hosted on the site. A description of the PhysioMIMI tools link was provided as an example of a resource hosted on an external site. Dan G. questioned if the ability to view and process physiological signals in the cloud would be available versus having to download the files to a local machine. Ary G. stated this could be possible in the future, but deferred to Remo who explained the current plan is to provide small tools for download that can be extended by the user or a programmer. Susan R. suggested, as a topic for a future call, the group consider what functionality they would want if a resource existed to allow access to data easily.
  5. Regulatory Issues Susan R. explained that NSRR is trying to be as congruent as possible with BioLincc and to closely collaborate with that group. BioLincc has agreed to provide a link on their site to when the BioLincc data repository contains data from research participants who also have PSG data residing in the NSRR. NSRR has agreed to do the same. In the future, NSRR hopes to form a similar relationship with dbGaP. Susan R. also discussed the need for a Data Use Agreement (DUA) and IRB review at sites wishing access to NSRR data. Susan R. will work with cohorts to ensure Steering Committees are comfortable with the data that is being shared.

  6. Questions from the AUG

    • a. Simon W. requested a high level overview of the Sleep-Terms group and asked about how the group plans to deal with variables, especially for longitudinal data. Susan R. suggested a 1 hour webinar devoted to this topic. A high-level overview of the canonical data dictionary was provided. Remo described the data dictionary foundation underlying the current process. The Sleep-Terms group is working on a mapping matrix.
    • b. Paul P. hopes to make the Wisconsin cohort data available and ask for key contact information. A contact list will be provided, and Susan R. suggested keeping contact information online as well as possibly using the Wisconsin data as as a use case for how to bring a cohort or dataset online in NSRR.

Action items:

  • Distribute slide presentation used for meeting in PDF format (Dan M.)
  • Create a list of key NSRR contacts for AUG (Dan M., Remo M.)
  • Arrange future calls, when appropriate, to discuss
    • 1) desired functionality of a resource that allows easy access to data and
    • 2) canonical data dictionary and how variables are handled.
Write a Reply