The document discusses the formation and goals of the Global Alliance for Genomics and Health. It was started in 2013 to facilitate international sharing of genomic and clinical data. Its goals are to establish common frameworks for data sharing, catalyze specific data sharing projects, and demonstrate the value of aggregating data from many sources. It currently has over 200 partner organizations from 30 countries. Working groups are advancing priorities around regulatory issues, data standards, security, and clinical implementation. The alliance aims to create a growing, sustainable network that continuously improves understanding of human health through large-scale data sharing and analysis.

1. Aarno Palotie
Director of the Human Genomics
Programme,
Institute for Molecular Medicine (FIMM),
Finland
Perspective from a Global Alliance

2. Aarno Palotie M.D., Ph.D.

3. The opportunity
An explosion of genomic information from individuals with
known clinical characteristics and disease outcomes
Learning from these data, we should accelerate progress in:
• Understanding the basis of inherited disease
• Cancer outcomes and targeted therapy
• Identifying targets for drug development
• Infectious disease
Clinical interpretation of individual genome sequences

4. The challenge: we can write down genomes, but we don’t yet know how to read them

5. To learn, we must compare
vs
vs
Cancer vs Normal

6. The challenge
Data from millions of samples may be needed to achieve
results and progress - showing patterns that would
otherwise remain obscure.

7. Example: common diseases, common variants

8. 0
5
10
15
20
25
30
35
40
45
Jan-10 Jan-11 Jan-12 Jan-13 Jan-14
Loci associated with migraine at genomewide significance
Migraine with aura
2,731 cases vs.
10,747 controls
5,122 cases vs.
18,108 controls
Migraine without aura
2,455 cases vs. 4,611
controls
Migraine GWAs
23,285 cases vs.
95,425 controls
58,930 cases vs.
314,166 controls

9. Example: rare mutations in families

10. Progressive Myoclonus Epilepsy (PME)
• Rare disease
• 84 samples collected around the world
• A recurrent de novo missense mutation in a
potassium ion channel gene (KCNC1)
• 13 unrelated patients out of the 84 patients (13.1%)
10

11. Deciphering Developmental Disorders (DDD)
Rare diseases
• 24 regional offices
• 12 000 children without a diagnosis,
including parents
• Whole exome sequencing

12. Example: ethnic diversity

13. Rare and low frequency vs. common variants
13
Single rare variant dominating the risk (>5×risk)
Multiple genetic and other risk factors contributing (>2×risk)
Courtesy of Emmi Tikkanen

14. Deletion of the TOP3B gene is a schizophrenia risk
Internal
Isolate
High risk for
schizophrenia

15. 4 individuals with homozygous deletion on 22q11.22
Stoll and Pietilainen et al Nature Neuroscience 2013

16. Example: new drug targets, Protective Loss of Function Mutations

17. Two LPA LoF variants protect
for Coronary Heart Disease
LoF variant Frequency in Finns Frequency in non Finns
LPA1(4974) 2.8% 0.47%
LPA2(4289) 4.8% 3.6%
227 Finns LoF homozygotes
Linking to National Health Records
No sign of increased morbidity
35 000 Finns tested in a population cohort

20. Sequencing 100 000 genomes
GWAs of 500 000 samples

21. The challenge
Data from millions of samples may be needed to achieve
results and progress - showing patterns that would
otherwise remain obscure.
That will take new methods and organizational models.
• Data is typically in silos: by type, by disease, by institution
• Analysis methods are non-standardized, few at scale
• Approaches to regulation, consent and data sharing limit
interoperability
If we don’t act: risk a hodge-podge of Balkanized data, such
as electronic medical records in the USA
Create interphases: API, create a Web

22. What to do?
Work together internationally to ensure interoperability of
data and of methods, to harmonize approaches to ethics
and regulation, and to promote participant autonomy
Support pilot projects that responsibly and effectively
harmonize, analyze and share genomic and clinical data
Demonstrate the value of aggregating and sharing data
Engage professional communities and the public to build
trust and encourage appropriate sharing and learning

23. Global Alliance in 2013
Starting in 2013
January 2013: 50 people from eight countries met in
NYC to define the problem and consider solutions
June 2013: after having engaged 80 people in writing a
White Paper, we announced the formation of the
Alliance with 70 organizations as Partners to take on
the challenge
December 2013: four Working Groups up and running;
Expanded Steering Committee; Executive Staff at Host
Organizations; Progress on governance, branding…

24. Mission
To accelerate progress in human health by
helping to establish a common framework of
harmonized approaches to enable effective
and responsible sharing of genomic and
clinical data, and by catalyzing data sharing
projects that drive and demonstrate the value
of data sharing

25. Theory of change
After discussion and feedback, it was concluded that the
Global Alliance has the best chance to succeed and
achieve our mission if we:
• Gather a broad and diverse network of stakeholders
• Publicly commit to advancing progress in data sharing
• Establish common frameworks of approaches
• Catalyze interactions and shared activity among the
members: in particular data sharing pilot projects that
drive the work of and are supported by the Alliance

26. Roles we can play
• Convene stakeholders to share information and methods
• Catalyze data sharing among members, support pilot projects
• Identify best practices and where needed develop new ones
• Act as clearinghouse, create network effect by sharing data
and methods, cross-pollinating ideas, and communicating
• Foster innovation by identifying driving problems, bringing
together experts, lowering barriers to new methods
• Commit to responsible data sharing, working together to
promote high standards for ethics and ensuring participants have
the choice to share data if they so choose

28. 2014
In first four months of 2014
Four initial working groups advancing their topics
First face-to-face meeting of Alliance partners at the
Wellcome Trust on March 4th: included 180 participants
from 100 partner organizations and 17 countries, with four
day-long satellite meetings (one on each working group)
Membership has grown to 225 partners in 30 countries,
including leading information and life science companies

29. Partner overview
225
Partner Organizations include:
Universities and Research Institutes
Academic Medical Centers and Health Systems
Disease Advocacy Organizations and Patient Groups
Consortia and Professional societies
Funders and Agencies
Life Science and Information Technology Companies
Last Update: April 16th, 2014

30. Partner overview
From A-Z (a very partial list)
Last Update: April 16th, 2014
Amazon Web Services
American Association for Cancer Research
American College of Medical Genetics
American Society of Clinical Oncology
American Society of Human Genetics
Amgen
Appistry
Association for Molecular Pathology
Australian Genome Research Facility
BBMRI-Netherlands
Belgian Medical Genomics Initiative
BGI-Shenzhen
Biogen Idec
Boston Children’s Hospital
Brigham and Women's Hospital
Broad Institute of MIT and Harvard
Sanofi
SAP Labs
Seven Bridges Genomics
SIB-Swiss Institute of Bioinformatics
Simons Foundation
Spanish National Cancer Research Center
St. Jude Children's Research Hospital
The Jackson Laboratory
The Personal Genome Project
THL, Finland
U of Texas M.D. Anderson Cancer Center
University of Toronto
University of Washington
Vanderbilt University Medical Center
Wellcome Trust
Wellcome Trust Sanger Institute

31. Partner overview
Last Update: April 16th, 2014
30
Countries in which alliance partners are based:
Argentina Australia Austria
Belgium Brazil Canada
China Finland France
Germany Hungary India
Ireland Japan Mexico
Netherlands New Zealand Singapore
Spain South Africa Sri Lanka
Sweden Switzerland United Kingdom United States

32. Lean, distributed operation
Staff at multiple Host Institutions (currently three)
• Ontario Institute for Cancer Research
• Wellcome Trust Sanger Institute
• Broad Institute / Brigham and Women’s Hospital
Transitional Steering Committee, 12 Members
Four Initial Working Groups to advance specific topics
• Regulatory and Ethics
• Data
• Security
• Clinical

33. Steering Committee
David Altshuler, Chair
Broad Institute, MGH
Martin Bobrow,
Vice Chair
University of Cambridge
Kathryn North,
Vice Chair
Murdoch Childrens
Research Institute
Paul Flicek
European Bioinformatics
Institute
David Haussler
University of California,
Santa Cruz
Thomas J Hudson
Ontario Institute for
Cancer Research
Kazuto Kato
Osaka University
Bartha Maria Knoppers
McGill University
Elizabeth Nabel
Brigham and Women’s
Hospital
Brad Margus
Genome Bridge
Charles L Sawyers
Memorial Sloan Kettering
Cancer Center
Peter C. Goodhand.
Acting Executive
Director
Global Alliance for
Genomics and Health

34. How we Work
• Working Groups with committed expert membership
develop initial work plan and priorities, supported by staff
• Creating Task Teams and Project Teams to develop
work products to address priority needs
• Distribute initial drafts for comments and input
• Take comments on-board, iterate and finalize
• Reference implementation and adoption

35. Initial Working Group chairs
Regulatory and Ethics
Bartha Knoppers
Partha Majumder
Kazuto Kato
Data
David Haussler
Richard Durbin
Security and Privacy
Paul Flicek
Dixie Baker
Clinical
Charles Sawyers
Kathryn North

36. Working Group current Priorities
Regulatory and Ethics: Framework for Genomic and Clinical
Data Sharing; Preparation of generic consent clauses; Ethics
review Safe Harbor and Data safe havens; Data Protection
Regulation.
Data: Governance of and updates to standard file formats
(BAM, CRFAM, VCF); Development (API (currently v.0.5) to
enable exchange of variation in genome sequences across
groups
Security and Privacy: identification and dissemination of
technical standards for secure exchange; identify privacy needs
Clinical: Assessment of methods for disease phenotyping;
matchmaking toolbox to enable gene discovery
Visit our website for an updated: Priorities Document

37. Working Group current Priorities
Regulatory and Ethics: Bartha Knoppers,
Participating in research is a human right
Data: Governance of and updates to standard file formats (BAM,
CRFAM, VCF); Development (API (currently v.0.5) to enable
exchange of variation in genome sequences across groups
Security and Privacy: identification and dissemination of technical
standards for secure exchange; identify privacy needs
Clinical: 1. Matchmaker project. You find a mutation -> have
someone else
2. BRCA challenge -> BRCA mutation catalogue
Visit our website for an updated: Priorities Document

38. Long-term vision, near-term goals
Long-term
• A growing network that engages the community
• A learning system in which data and models
continuously improve
• The potential to incorporate other types of data
Near-term
• Demonstrating value
• Set and achieve practical working group goals
• Organize around high priority projects
• Define an effective sustainable model
• Increase global engagement
• Next meeting October 18 in San Diego

39. Website: http://genomicsandhealth.org