From Charge
Jump to: navigation, search

Cohorts for Heart and Aging Research in Genome Epidemiology (CHARGE) Consortium Aggregate Results Sharing Principles

Approved by CHARGE Research Steering Committee on 5/28/2008

Overall Guiding Principle:

A major goal of the CHARGE Research Consortium is to collaboratively produce a series of jointly coordinated, high-impact publications that describes the collaborative results of genome-wide association (GWA) scans for a number of cardiovascular, lung, blood and aging phenotypes. The focal point of the design, analysis, interpretation and publication of research results for each specific phenotype is one of several collaborative Phenotype Working Groups (PWGs) that have been convened by the CHARGE Research Collaboration Committee. One analytic strategy that is being undertaken by many of the CHARGE PWGs is to conduct a meta-analysis of GWA results for single phenotypes common to multiple CHARGE cohorts using a set of HapMap SNPs imputed from the original genotyping platform. The actions of members of PWGs from each of the individual CHARGE cohorts and other collaborating cohorts should support this research goal. The principles in this document are meant to encourage collaborating investigators to conduct their research in the spirit of collaboration and trust and with transparency regarding potentially sensitive issues about the use of shared data prior to publication.

These principles are in no way meant to preclude each participating cohort from freely publishing their own study specific data. However, the prior publication of genome-wide data on a particular phenotype by a specific cohort may limit the ability of that cohort to participate in the initial consortium-wide meta-analysis of data for that same phenotype. This decision is best made among members of the individual PWG. However, each cohort is encouraged to clearly communicate its plans regarding participation in data sharing versus separate publication clearly to other members of the PWG and the CHARGE Steering Committee (RCC).

Timing of Opt-in for Specific Manuscripts by Members of Phenotype Working Groups

  1. Members of each PWG will collaboratively propose one or more specific manuscripts to be published making use of GWA scan data from one or more participating CHARGE cohorts as well as other collaborating cohorts invited by the PWG.
  2. Each cohort participating in a PWG will designate an investigator who is the primary leader of his or her cohort for each specific proposed manuscript.
  3. The lead investigator for the relevant PWG from each CHARGE cohort and from each additional collaborating cohort may decide that his/her cohort will “opt in” to participate in each specific manuscript that is proposed by the PWG. By “opting in” to participate in the PWG manuscript proposal, the analysis planning and authorship decisions will proceed under the assumption that all proposed investigators from each cohort will participate in the write-up of the manuscript. Also, the decision to “opt in” represents a commitment to collaborate only with the CHARGE PWG for the analysis and follow up of GWAS results described in the manuscript proposal, until the initial proposed PWG manuscript is accepted for publication or unless there is a consensus within the CHARGE PWG about a second collaboration.
  4. The proposal for collaboration leading to a specific manuscript will be accompanied with an agreed upon timeline and specific target dates for each cohort to complete GWA analyses by each participating cohort, for sharing of aggregate GWA analyses, for completion of any additional analyses or genotyping, and for submission of the manuscript to a journal.
  5. Sharing of aggregate GWA analyses by each participating cohort with available data will be necessary to move forward each collaborative manuscript. In any given PWG, some participating cohorts may complete their GWA analyses and, as anticipated in the planned timeline, be ready to share results in advance of other participating cohorts. In order to proceed with timely publication of GWA results, PWGs may determine that sharing and preliminary meta-analysis of aggregate GWA analyses take place by two or more participating cohorts before all participating cohorts are ready to share. Other investigators from other cohorts who have opted in will participate fully in review of such shared results and the authorship of the proposed manuscript(s), as previously agreed.
  6. All investigators who have “opted in” to participate in the proposed PWG manuscript will be granted access to shared results, in a manner determined by the designated lead investigator for each cohort, as described below.
  7. After the first occurrence of formal sharing of aggregate research results between two or more participating cohorts and the distribution of these shared results to participating investigators, no investigator who has “opted in” may then “opt out” of participation in the planned consortium manuscript in order to publish their own results separately or to publish collaborative results with other cohorts/consortia, until the aforementioned manuscript is published. At any time before formal distribution of shared results to all participants of the PWG, investigators have the option to opt out. When GWA data used in a CHARGE PWG manuscript are also proposed for use in other collaborations (e.g., other meta-analyses), it is expected that such use will be discussed in the PWG, and a timeline for publication of data in such additional collaborations will be agreed by consensus by collaborators in the original CHARGE PWG manuscript. The decision to opt in for participation in a manuscript should be made after the working group has agreed to an overall timeline for completion of the manuscript from the proposed meta-analysis.

Primary Results Distribution (Prior to Publication):

  1. Aggregate results files formatted according to the recommendations of the CHARGE Analysis Subcommittee and the members of the PWG will be submitted for a formal meta-analysis available to each cohort that is providing research results to all participating PWG investigators/cohorts via a previously agreed upon secure FTP site. Each planned occurrence of sharing will occur by all agreed upon cohorts within a prespecified 72 hour period. The shared files will include a complete listing of aggregate association results for >2 million imputed SNPs and appropriate documentation and may also include an abbreviated listing for all SNPs with association P<0.001. The principles described herein refer to aggregate genotype phenotype results, not participant specific data. Specific recommendations for sharing of primary participant-specific data from one or more cohorts participating in the CHARGE Consortium may require detailed discussion, given sensitivities regarding privacy and confidentiality of participant-specific data and will be made separately from the current document.
  2. Each PWG will identify a specific investigator(s) who will conduct the primary meta-analysis of shared data using a mutually agreed upon method in advance of the first formal sharing of aggregate results. The designated lead investigator from each cohort or his/her designee may download the aggregate results files for use in his or her own analysis. This lead investigator will take responsibility for sharing of these aggregate data with other collaborators from his/her cohort.
  3. Individual investigators from the participating cohorts who have opted in to participate in a PWG analysis/manuscript may use the shared aggregate results files to compare with other datasets that may not be available to all groups. An individual investigator who has opted in to participate in the PWG analysis and who has access to other data may use these additional data to further guide their experiments. However, the individual investigator(s) agree(s) to communicate with the rest of CHARGE PWG and CHARGE RCC about any actionable results, as described below. The principle here is that interesting results that emerge from a combination of external and CHARGE meta-analysis data will be treated the same way as if they arose entirely from CHARGE PWG data. However, secondary distribution of unpublished CHARGE PWG aggregate results to investigators or other persons who have not explicitly “opted in” to the relevant CHARGE PWG may occur only after approval by the PWG and confirmation by the CHARGE RCC. In general, such secondary distribution will be discouraged until after the relevant manuscript has been accepted for publication.
  4. Communication. All individual PWG investigators agree to keep the other investigators in the CHARGE PWG informed about additional analyses and follow-up experiment plans. Specifically, participating groups agree at a minimum to communicate with each other at the following times:
    1. When action is taken by an individual CHARGE investigator (for example, ordering primers for genotyping) based on shared lists of results.
    2. When an analysis is complete (with some degree of confidence) for follow-up genotyping from shared lists of results.
  5. In line with the emerging standards in the field and the requirements from journals, we expect that the entire CHARGE association aggregate results file as well as cohort-specific aggregate GWAS results files for a PWG endpoint analysis will be made publicly available upon publication.
  6. Any sharing of aggregate results not stipulated in this principles document is not allowed unless the issue is brought to the CHARGE RCC Steering Committee for discussion and approval.

CHARGE Publication & Presentation (P&P) Proposals:

  1. Each PWG proposal will be briefly described in writing in a CHARGE Publication & Presentation (P&P) Proposal to be submitted to the RCC and posted on the CHARGE Wiki.
  2. The CHARGE P&P Proposal will include a description of the proposed analysis, specific phenotypes for analysis, names of lead investigator(s) and other cohort investigators opting in to participate from each cohort, specific authorship order including starred authors, and the proposed timeline for completion of analyses a manuscript.
  3. The CHARGE P&P Proposal will further include specific target dates for each cohort to complete GWA analyses by each participating cohort, for sharing of aggregate GWA analyses, for completion of any additional analyses or genotyping, and for submission of the manuscript to a journal.
  4. Submission of preliminary results of any given PWG to be presented as an abstract at a scientific meeting or other pre-publication presentation of results including presentation of pre-publication results in grant applications must be agreed upon by all members of the PWG and must be consistent with the initial P&P Proposal.

Signatures of Lead Investigator from Each Participating Cohort

Signature Date
ARIC -- Eric Boerwinkle
AGES -- Tamara Harris
CHS -- Bruce Psaty
FHS-- Christopher O’Donnell
Rotterdam -- Albert Hoffman