MEMORANDUM OF UNDERSTANDING

29 March 2017

This document seeks to outline the terms of reference for the oversight, governance, activity and output of the HERMES Consortium.

Aim HERMES is an international research collaboration to explore the genetic determinants of heart failure risk and prognosis in order to identify and validate therapeutic targets and causal risk factors.

Ethos The scientific strength of HERMES lies in its collaborators. The guiding principles for HERMES are scientific excellence; collaborative spirit; transparency and openness, ‘no surprises’; equal opportunities for contribution across cohorts; open communication; and mutual trust and confidentiality. The aim of this collegial approach is to maximize the scientific potential of the consortium and to ensure that all members receive appropriate recognition in the publications that will be generated. The consortium will be inclusive and, as such, open to collaboration with both academic, industrial and charitable partners who agree to participate according to the terms of this memorandum.

Organisation The management of HERMES will be divided amongst teams: scientific leadership and oversight will be provided by the Scientific Committee (SC); general management and logistic support will be provided by Executive Committee (EC); and primary research projects will be delivered by project working groups. Further detail on each of these groups is given below.

Scientific Committee The SC will comprise representatives from each of the participating cohorts and invited members with particular expertise in relevant disease areas or research methodologies. It is envisaged that a senior investigator from each of the cohorts will join the SC. The SC will provide scientific leadership and overall oversight for HERMES Consortium and HERMES Projects. The SC will meet at approximately two-monthly intervals ‘all-hands meeting’ (with the option of more frequent meetings). Quorum will be set at 30%. Key roles will include:

Executive Committee The EC will form a subgroup of the SC and will be responsible for providing administrative support and project management for the consortium (Appendix 1). Key roles include:

Project Working Groups The primary scientific work of the HERMES consortium will take place in working groups around specific meta-analysis projects. Any consortium member may propose a new project and working group at an SC all hands meeting. Each working group will consist of 1-2 coordinator(s) who will be responsible for the design, analysis and reporting of the project, as well as day-to-day operational management of the project. In addition, working group will include 1-2 representatives from participating cohorts and a nominated lead analyst(s). Key roles will include:

Data sharing policies Cohort PIs retain all rights to their data. Only summary statistics are shared for meta-analysis. Aggregate (meta-analysis) data may not be used for other purposes than the working group aims without permission from the working group or the specific study, respectively. Sharing of unpublished results to non-members of the working group that have not explicitly opted-in to the analysis may only occur after approval by the working group. This may, for example, be desirable in situations where a working group is seeking lookup-replication, statistical advice, or functional analyses by groups that are not participating in HERMES.

For HERMES GWAS meta-analysis, individual study results will be transferred to the HERMES analysis centre for QC and meta-analysis (currently University College London), and then shared with a second analysis centre for meta-analysis replication (currently Boston University). Individual study results will not be used for any other purpose and will be treated as confidential by the HERMES analysis centres. Upon completion of the meta-analysis, results will be made available to HERMES Contributing Members (defined below) for review and then made publically available online upon publication of the relevant results paper.

  1. Data freeze for studies contributing to HERMES meta-analysis (week -5)
  2. Completion of QC and meta-analysis replication (week 0)
  3. Confidential release of summary-level data to HERMES Contributing Members only (week +1)
  4. Summary-level data made publically available online upon publication of relevant results paper  (approx. week 36 - 52)

For the purposes defining rights to pre-publication data access, HERMES members are classified as follows:

HERMES Contributing Member

Definition: Cohorts/ organisations who have agreed to the MOU and have contributed genome- wide summary statistics to a HERMES GWAS project either directly, or indirectly by supporting another participating cohort (e.g. genotyping or analyst support), or who have provided project funding to the HERMES core analysis centres.

Data access: Contributing members will have equal access to the results of HERMES GWAS meta-analyses according to the internal data-sharing policy (1 week after final meta-analysis results generated). Contributing data to one analysis will grant access to all subsequent GWAS and sub group analyses. Genome-wide meta-analysis estimates will be downloadable from a password protected site managed by the coordinating centre.

HERMES Affiliate Member

Definition: Cohorts/ organisations that have agreed to the MOU and have expressed interest in contributing data to future HERMES analyses (when data/ analysis capacity permits).

Data access: Participation in SC meetings to steer scientific strategy. SC meetings may present results in tables and figures for discussion. Affiliate members will not be able to download GWAS meta-analysis estimates.

Opt-in and opt-out policy Investigators may join each working group or may opt out of one or more working groups for any reason. The decision to ‘opt in’ represents a commitment to collaborate only with the HERMES working group for that particular analysis, until the initial proposed manuscript is accepted for publication. This includes a commitment not to accept new replication requests for the topic under study from researchers not participating in the HERMES working group. No restrictions will be put upon publishing the data that an individual study would contribute to the meta-analysis before opting in to the meta-analysis. Uploading data for meta-analysis will be assumed to imply opting-in to that specific project. At the time of opting in, researchers are required to disclose potentially conflicting ongoing work on the topic of the project (‘no surprises’ ethos). The decision to opt out of an analysis must take place before the first occurrence of sharing of any research results. After the results have been shared, investigators cannot opt out to publish their findings on their own or with other partners. Only investigators from studies that have opted-in have access to shared results and will be invited to teleconferences in which these results are discussed.

Industrial collaborators HERMES will seek to enrol clinical trials cohorts and as such welcomes the participation of industrial collaborators. Industrial and academic collaborators who contribute cohorts to the consortium will participate on an equal basis to the consortium. All results from HERMES analyses will be made publically available in due course through online repositories and in scientific publications.

Publication planning An initial publication plan will be developed at the time of working group formation and may be modified based on the results following agreement from the SC. Publications should be prepared and submitted as rapidly as possible after the results have been reviewed and confirmed. For parallel publications, one month is a tolerable wait time; three months may not be tolerable. Each working group needs to balance rapid publication, strength of findings from replication, and equal partnership. Working group members will assure that parent-study disclaimers, reviews, and approvals are managed as required to avoid any delay in submission. These approvals should occur within one month of the completion of the final draft. Abstracts should follow the same rules as publications.

Authorship The goal is fair scientific representation from cohort members participating in the working group. Multiple first and multiple senior authors may be designated as such.

First and senior authors, including the overall first and the overall last author, should generally come from different cohorts. Long lists of authors are permitted if all authors meet standards for authorship (such as those required by major journals, for instance the criteria used by Nature Genetics and JAMA). Contributing scientists from outside the working group may be co-authors, however, in the event that two people propose similar projects at the same time priority will be given to the internal team. Authorship position in first papers will be determined in advance by a group consensus by the researchers actively participating in the proposed pooled study. Some of the factors that may be considered include effort in working group and size of participating cohorts study. In later papers, successful efforts to obtain additional funding for new scientific work is another criterion. The SC will mediate any differences of opinion regarding authorship representation.

Rapid publication from a working group should generally include authors from all member cohorts contributing summary statistics from the specified analysis. The number of authors per study should generally reflect the contribution of the study to the paper and will be flexible and determined on a case by case basis. Consortium authorship for the writing group is another possible model that may be considered as an alternative model if there is no group consensus on named authorship.

Involvement of Other Collaborating Studies Other collaborating studies may be proposed for inclusion in individual working group analyses. Collaborating studies will be asked to abide by all of the principles set out in this memorandum for the purpose of the proposed working group project.

Data posting For the purposes of transparency and quality control, meta-analyses results of will be made available to consortium members through HERMES Dataview as soon as they are available following appropriate quality control processes. Following publication in a scientific journal, shared summary-level results including results from joint analyses will be posted on www.hermesconsortium.org for the benefit of the wider community, under terms of use that will be agreed by the SC in advance of posting.

Intellectual property Concerning meta-analyses performed within the HERMES Consortium (GWAS, pathway analyses and Mendelian randomisation studies) the principles outlined for NIH funded GWAS studies (Federal Register; 72 (166) 28 Aug 2007: 49290-7) will be followed: “genotype- phenotype associations identified through NIH-supported and NIH-maintained GWAS datasets and their obvious implications will remain available to all investigators, unencumbered by intellectual property claims.” In the event that intellectual property related to the GWA studies covered under these principles is pursued by a participating investigator, their institution, or the study itself, the SC must be promptly made aware of the pursuit of such claims. An abstract of the claim(s) needs to be made available to the consortium at the time of submission to the relevant governing body (e.g. US Patent Office).

Acknowledgements HERMES thanks iGeneTrain, GENIUS-CHD, CHARGE and PACE consortia for sharing their memoranda of understanding upon which this document is based.


Appendix 1

Executive committee members (as of 12 October 2016)

Tom Lumbers

University College London

(Coordinator)

Dan Swerdlow

University College London

(Coordinator)

Folkert Asselbergs

UMC Utrecht

Aroon Hingorani

University College London

Chim Lang

University of Dundee

Vasan Ramachandran

Boston University

Gustav Smith

Lund University

Jemma Wilk

Pfizer