Technology to Enable the Genomics Revolution How Melbourne Genomics is Delivering for Patient Care Kate Birch Data & Technology Program Manager
Topics for discussion • Delivering a shared platform for bioinformatic collaboration as part of GenoVic • Storage of clinical genomic data as part of GenoVic Melbourne Genomics Health Alliance | Document Name Here 2
From genetics to genomics • Genetics scrutinizes the functioning and composition of the single gene where as • Genomics addresses all genes and their inter relationships in order to identify their combined influence on the growth and development of the organism World Health Organisation Melbourne Genomics Health Alliance | Document Name Here 3
From genetics to genomics Integration with microbiome, proteomics, metaboloimics... Whole genome Whole exome Large panels Small panels Single gene Melbourne Genomics Health Alliance | Document Name Here 4
From genetics to genomics Integration with microbiome, proteomics, metaboloimics... Whole genome Whole exome Large panels Small panels Single gene Melbourne Genomics Health Alliance | Document Name Here 5
Melbourne Genomics Health Alliance 6
An analogy….. 1000 copies of War and Peace = a single genome
An analogy….. Shred them. Read each piece and reconstruct the story. Find the typos. 1000 copies of War and Peace = a single genome
An analogy….. Shred them. Read each piece and reconstruct the story. Find the typos. 1000 copies of War and Do they change the meaning of the Peace = a single sentence? genome
Research Clinical care Melbourne Genomics Health Alliance | Document Name Here 10
Only 2% of US genomic data is used for clinical care Erik Jylling - Executive Vice President Health Politics, Danish Regions 11
The Melbourne Genomic Health Alliance set out to make Victoria a world leader in the translation and use of clinical genomics in healthcare 12
Challenge: Create whole of system change 13
Melbourne Genomics Health Alliance Melbourne Genomics Health Alliance | Document Name Here 14
Melbourne Genomics Health Alliance Melbourne Genomics Health Alliance 15
Approach STREAM 3: INNOVATION AND RAPID ADOPTION Develop and deploy systems to ensure patients STREAM 1: have access to cutting-edge, high quality genomic WORKFORCE DEVELOPMENT testing that is cost-effective. STREAM 5: NATIONAL AND Build the literacy, skills and INTERNATIONAL IMPACT confidence of the clinical and diagnostic workforce in Establish active relationships and participation in genomics, as relevant to each national and international initiatives with the aim professional role of disseminating, communicating and collaborating on the work of the Alliance and its implications. DISEASE FLAGSHIPS Flagships are the mechanism through which genomic sequencing is provided to patients with defined clinical conditions or indications. STREAM 2: ASSESSING THE VALUE Flagships will also be the means by which the OF GENOMICS workforce is developed, innovation is adopted, outcomes are evaluated and information systems Evaluate the place of genomics in health care trialled, and underpin the five streams. practice, by: (1) evaluating the process and outcomes of genomic tests in practice, and STREAM 4: ACCESS TO GENOMIC INFORMATION (2) establishing and applying a platform for Develop and implement a single set of standards, policies health service research, program evaluation, and procedures to support a common infrastructure for the economic evaluation and translational research management and use of genomic data by stakeholders in in the use of genomics in health care. Victoria. . . Melbourne Genomics Health Alliance 16
Technology to enable clinical genomics 17
Access to Genomic Information Develop and implement a single set of standards, policies and procedures to support a common infrastructure for the management and use of genomic data by stakeholders in Victoria. Melbourne Genomics Health Alliance 18
Ideal end state in an unconstrained environment People Technology Policy & Process 7. Identity & Access Management 1. Standardised policy and 8. Clinical Tools 10. Patient Tools 9. Diagnostic Tools processes for data 5. Staff to management & access Electronic manage the Clinician Consent (data governance) Analysis Orders and data Knowledge (Pipeline) Tools Results Results Clinical Decision Support Curation Tools Education Tools 2. Standardised policy & processes for patient consent 11. Data Access Tools 6. Staff to manage the technology 12. Master 13. Genomic Data Repository 3. Standardised policy and Patient Index processes for test ordering & reporting 14. Data Integration LIMS 4. Change control process EMR Public variant (genomic (clinical data) curation data sequencing data) Melbourne Genomics Health Alliance 19
Diagnostic Tools Approach LOVD EOI Prototype Future Workshops Cpipe learnings needs Pilot SeqLiner Prototype Define Requirements Procure Melbourne Genomics Health Alliance 21
Melbourne Genomics Health Alliance 22
GenoVic G S Genomic Orchestration Service
A shared platform for bioinformatic collaboration 24
GenoVic G S Genomic Orchestration Service
Bioinformatics Challenges Pooling of resources, collaboration Clinically hardened software Clinical grade compute Scalable, cost effective compute 26
DNAnexus A platform for developing and running bioinformatics pipelines on AWS 27
Bioinformatics Challenges Pooling of resources, collaboration Clinically hardened software Clinical grade compute Scalable, cost effective compute 28
Pipeline Design Each app accepts inputs from any number of samples • Solves code duplication • Supports batching Each app uses a Docker image to run the analysis • This keeps our workflows portable • Allows us to leverage existing containers such as the miniconda environment 29
Acceptance criteria outcomes - runtime Runtime - trio Runtime for a Ashkenazim Trio child (NA24385) is 12h 29m. 30
Framework for testing shared pipelines The Alliance has implemented a basic framework which the laboratories can build on to implement the comprehensive pipeline testing required for accreditation. 31
Bioinformatics Challenges Pooling of resources, collaboration Clinically hardened software Clinical grade compute Scalable, cost effective compute 32
Variant identification Melbourne Genomics Health Alliance | Document Name Here 33
DNAnexus Germline Pipeline Alliance Pipeline V1 New Rapid Alliance Member's Own reference exomes? Pipeline V2 Pipeline genome? Long reads? Routine QC? Alliance CWL/WDL? Pipeline V3 Alt aware? 34
Bioinformatics Challenges Pooling of resources, collaboration Clinically hardened software Clinical grade compute Scaleable, cost effective compute 35
Storage of clinical genomic data 36
Why store clinical genomic data? Pathology labs need to Reanalysis leads to new diagnoses Clinical reference data Valuable research data 37
GenoVic G S Genomic Orchestration Service
Storage of genomic data Store Catalogue & search Release 39
Storage of genomic data Store Catalogue & search Release 40
Storage of genomic data Store Catalogue & search Release Policies and procedures for data sharing Data Governance Owners, Stewards and Alliance Data Custodians 41
Patient’s views on data sharing Melbourne Genomics Health Alliance | Document Name Here 42
43
44
We asked patients having genomic testing what they want 45
Patient accept genomic testing More than 96% of patients consented Melbourne Genomics Health Alliance
Common clinical consent form Sharing of anonymised data Opt in to share data for activities related to their condition Melbourne Genomics Health Alliance
Majority are informed and not concerned Patients are informed after genetic counselling • Over 80% of patients correctly answered questions about the types of results they may receive and storage of their data Almost everyone agrees to additional use of re-identifiable data • 99% agree to additional use of re-identifiable genomic data when giving clinical consent Majority have no remaining concerns about sharing their data • 97% received enough information about data sharing • 93% have no remaining concerns about sharing data • 13 had concerns (privacy, potential for discrimination – employment, insurance) Melbourne Genomics Health Alliance
Most not worried even if they are identified Ease of identification Easy Difficult Difficult Easy & & & & & Level of concern if concerned concerned Unconcerned Unconcerned identified Melbourne Genomics Health Alliance 49
How broadly should data be shared? Alliance Members High trust Patient trust in sharing data Australian not-for-profit Overseas not-for-profit Pharmaceutical Government Other Industry Low trust Melbourne Genomics Health Alliance 50
Recommend
More recommend
