Download Quality assurance and guidelines for validation of next

Quality assurance and guidelines
for validation
of next-generation sequencing tools
Gert Matthijs
Center for Human Genetics
University of Leuven, Belgium
- on behalf of EuroGentest -
Genetic testing
Molecular testing
1 disease
1 (single) mutation
1 disease
1 gene
different diseases
1 (single) gene
1 disease
few genes
1 disease
1 chromosomal
anomaly
(un)known defect
karyotyping
different diseases
many genes
different diseases
different anomalies
total
genome
Molecular cytogenetic
testing
Cytogenetics
G. Matthijs 2011
Long QT syndrome ?
LQT1
LQT2
LQT3
LQT4
LQT5
LQT6
LQT7
LQT8
LQT9
LQT10
LQT11
LQT12
LQT13
ATTGTACGTGATGACCAGTGGAAT
ACCGTAAGGTAAAGTACCGTGTAC
TTGGTTGGAACGTAGACTGAATGC
CAACCCTGGTATTGGTGTCCCGTG
TACAAGGTTAGTAATGTACCATTGA
TTCCGTAATACGTGTGGCGCGTGC
GTAACACACTGACTGACCATCCTG
GTAGCTAGTCAAGTCGTAGCTGTC
GTGACGTAACGTATATGACACACG
TCAGTACGGTCAGTACACACATGC
TGTGGTGCAGTACAGATACAGTAC
AGATTAGCAGAAATGCAGATTTAGT
ATTGTACGTGATGACCAGTGGAAT
ACCGTAAGGTAAAGTACCGTGTAC
TTGGTTGGAACGTAGACTGAATGC
CAACCCTGGTATTGGTGTCCCGTG
TACAAGGTTAGTAATGTACCATTGA
TTCCGTAATACGTGTGGCGCGTGC
GTAACACACTGACTGACCATCCTG
GTAGCTAGTCAAGTCGTAGCTGTC
GTGACGTAACGTATATGACACACG
TCAGTACGGTCAGTACACACATGC
TGTGGTGCAGTACAGATACAGTAC
AGATTAGCAGAAATGCAGATTTAGT
ATTGTACGTGATGACCAGTGGAAT
ACCGTAAGGTAAAGTACCGTGTAC
TTGGTTGGAACGTAGACTGAATGC
CAACCCTGGTATTGGTGTCCCGTG
TACAAGGTTAGTAATGTACCATTGA
TTCCGTAATACGTGTGGCGCGTGC
GTAACACACTGACTGACCATCCTG
GTAGCTAGTCAAGTCGTAGCTGTC
GTGACGTAACGTATATGACACACG
TCAGTACGGTCAGTACACACATGC
TGTGGTGCAGTACAGATACAGTAC
AGATTAGCAGAAATGCAGATTTAGT
K. Devriendt 2012
Next Generation Sequencing (NGS) work flow
Targeted sequencing
Exome
Capture array
Library
Preparation
Sequencing
Variant
Annotation
Courtesy: Erika Souche 2013
Sequence
Capture
Demultiplexing
Variant
Filtering
Amplification
Mapping
Report
Variant
Detection
Illumina run quality control
Cluster density
Courtesy: Erika Souche 2013
Sample quality control
Base quality score by position
Courtesy: Erika Souche 2013
Mapping
Courtesy: Erika Souche 2013
Diagnostics quality control
Courtesy: Erika Souche 2013
Nat Biotechnology 20 (2012): 1033-36
Nex-StoCT (standardization of clinical testing)
Nat Biotechnology 20 (2012): 1033-36
Nat Biotechnology 20 (2012): 1033-36
Research
≠
Diagnostics
EuroGentest wants to build on these available documents and the
work of others, and then comment or amend where necessary.
* US: Nex-StoCT (CDC)
Nat Biotechnology 20 (2012): 1033-36
* UK: Targeted NGS Guidelines (CMGS)
* AUSTRALIA: Concept NGS Guidelines (MPS-WG)
* THE NETHERLANDS: Concept NGS Guidelines (VKGL)
* ESHG PPPC: Recommendations for WGS
Draft - Discussed at EuroGentest expert
meeting, February 2013
ISSUES of analytical VALIDATION
Cfr. US Guidelines
Platform verification: verify the performance specifications
established by the manufacturer (e.g., accuracy, precision, etc.
(FDA)
Test & Bioinformatics pipeline validation: validate system
performance characteristics (accuracy, precision, reportable
range, reference range, analytical sensitivity, analytical specificity,
and other performance characteristics, as applicable) for
laboratory-developed tests (LDTs) (CLIA)
Reportable range: areas of the targeted region that cannot be
sequenced reliably and therefore are excluded from the
reportable range
Draft - Discussed at EuroGentest expert
meeting, February 2013
ISSUES of DEVELOPMENT of the assays
Cfr. Dutch Guidelines
* Definition of diagnostic test
* Aim of NGS
* Analytical routing
* Core gene list
* Diagnostic yield
Draft - Discussed at EuroGentest expert
meeting, February 2013
DEFINITION OF A GENETIC TEST
Published evidence for a disease gene to be included in
NGS diagnostics (gene list)
This reference has to be included in the labs bibliography
Disease genes and mutations that are not amenable to NGS
testing (trinucleotide repeat expansions, deletions, …)
should not be tested by NGS only
(include that information to clinician)
Draft - Discussed at EuroGentest expert
meeting, February 2013
AIM OF NGS
- Is the test used to exclude a disease?
- Is the test used to confirm a disease?
Depending on this, the approach may differ, and also this will
affect how the results will be reported
Reporting of ‘uncertain variants’(UVs): no difference with the
way UVs will be reported for NGS as compared to current
practice
Draft - Discussed at EuroGentest expert
meeting, February 2013
CORE GENE LIST
A ‘core gene’ is a disease gene that has to be genotyped
completely because it contributes largely/importantly to the
disease prevalence.
* Other criteria may be: actionable, treatable…
* This may include gene dosage analysis
Core genes have to be outlined in the test description
Core gene should be outlined in BPG and in CUGC
Note: invite experts to generate those (minimal) lists
There is an economical aspect in these considerations
Draft - Discussed at EuroGentest expert
meeting, February 2013
REPORTING, REPORT TO THE CLINICIAN
Cfr. Australian guidelines
Technical details
Quality issues
Interpretation (strong preference to only report the known or
likely pathogenic variants)
* Requirements:
- figure showing the coverage
- list of genes that were analysed
* Re-contacting / Re-analysis
Draft - Discussed at EuroGentest expert
meeting, February 2013
DISTINCTION AND BORDERS BETWEEN
RESEARCH AND DIAGNOSTICS
Diagnostics is (published) evidence driven
Diagnostic test for genes when there is a known
relation between genotype and pathology
Research is hypothesis driven
Any further analysis should be subject to
Ethical Board Review
Question: where does evidence end?
The professional has to decide what is evidence and
what is not proven.
Draft - Discussed at EuroGentest expert
meeting, February 2013