Proposals for Additional Data Analysis

Secondary Analyses of Existing Summary Data

At the time of this writing (11/20/2013), all summary statistics will be protected on University of Michigan servers, and curated by Scott and Gonçalo. Proposals that call for additional analysis (i.e., nothing that GSCAN is already planning to do) of existing data will be discussed by local site investigators. Note that some investigators may have to obtain permission from committees overseeing their study before sharing summary statistics from that study (e.g., for very large and/or complex cohorts). If approved, these proposals will be granted access to selected portions of the summary statistics. Any manuscript generated from these secondary analyses must not be submitted for publication until the primary GSCAN manuscript has been accepted for publication.

Proposals that Require Site Investigators to Perform Additional Analysis

Proposals that request local sites to perform additional analysis will be more difficult to implement. In the end, such proposals will be "at the mercy" of local sites, who may or may not agree to conduct the proposed analyses. Such proposals can be discussed on conference calls and local sites may volunteer to participate.

Authorship for Publications Resulting from Proposals

The GSCAN consortium should be listed as an author, if possible, in any publication resulting from secondary analysis of GSCAN data. If listing a consortium name is not possible, for example due to journal guidelines, then individual GSCAN investigators may be listed. Further, depending on the level of involvement from GSCAN members in facilitating fulfillment of the proposal's aims, individual investigators may be named as well.

Data Security

Summary statistics of genetic variants are not truly de-identified, as there remains the possibility that they could still be used by a sufficiently sophisticated and motivated person to re-identify individuals. This concern must be taken very seriously, and any proposal to use GSCAN data, even if only summary statistics, must have a plan to assure the safety and security of the data, protect against its re-release, and must guarantee that there will be no attempt to re-identify the data.

Submitting Proposals

Please submit proposals to Scott Vrieze or Gonçalo Abecasis. Please include the following information:

  • Names of investigators, institutions, contact info, etc.
  • The data you need
  • Description of the experiment
  • If requiring additional analysis from local sites, an analysis plan
  • Timeline to completion of project