PF19: Genetic Question 7

For complex pancreatic diseases clinical research information sharing, which public databases shall be updated, and how shall information be shared through collaborative data sharing agreements?

Chairs: Ben Busby PhD, John Neoptolemos MD FRCS FMedSci,

Delegates: Péter Hegyi MD PhD, Andrea Párniczky MD 

Considerations:

What kind of new databases, or support structures are needed? New challenge: how to track and report exome and genome data? How to share data and harmonize data.

6 Replies to “PF19: Genetic Question 7”

  1. Prof Garg noted the approach of the GA4GH (see its website http://www.ga4gh.org). What does the group think of these guidelines related to data privacy and data sharing?

    For complex disorders we need all of the variable and phenotypes on individual patients – not just summary statistics.

    Also – what about the approach of the UK Biobank?

  2. In my personal opinion, the GA4GH DRS model will likely be a reasonable way to point to federated data resources.

    A question that I would like to propose to delegates and the public are about community incentivization of sharing data. What types of incentives would make putting data on front-facing servers or cloud infrastructure worthwhile?

  3. The NIH just came out with some new guidelines linked to dbGaP. This is the text of the email and the link to the PDFs.

    NIH streamlined the DUC agreement to reduce paperwork for projects that include requests for access to multiple datasets. Dataset-specific terms (e.g., Data Access Committee contact information, data use limitations, suggested acknowledgement statement) are now collated in an addendum appended to the end of a single copy of a standardized DUC agreement.
    Important updates to the terms described in the DUC agreement and Genomic Data User Code of Conduct are outlined below:
    · Both the DUC agreement and the Genomic Data User Code of Conduct now include a statement that Approved Users (e.g., Principal Investigators, collaborators, trainees) may not generate information (e.g., facial images or comparable representations) that could allow the identities of research participants to be readily ascertained.
    · In the DUC agreement, expectations about retention of data after close-out have been updated.
    · The DUC agreement outlines the legal obligations of Approved Users regarding data disclosure resulting from dbGaP’s Certificate of Confidentiality and subsection 301(d) of the Public Health Service Act.

    https://osp.od.nih.gov/wp-content/uploads/Model_DUC.pdf

    http://osp.od.nih.gov/wp-content/uploads/Genomic_Data_User_Code_of_Conduct.pdf

  4. Regardless, I think given cloud infrastructure, we could avoid this model altogether. If we had a computational contract based on something like CWL, which dictates exactly what kind of data is returned, there is no need for an individual to have full access to most datasets.

  5. Very important issue.
    – I think an IT development is necessary.
    – cloud infrastructure is important
    – one database should be built up. This would allow a uniform data collection and would make the analysis easy. All participants should keep the rights for all studies to share or retain their data.
    – GDPR alaso a very important issue which needs to be considered
    .

Leave a Reply