Dealing with Genetic Variation Data: A Diagnostic Perspective

1
Dealing with Genetic Variation Data A
Diagnostic Perspective

• Chris Mattocks
• National Genetics Reference Laboratory (Wessex)

2
Outline

• Diagnostic laboratories
• Data types usage
• How is data currently handled
• Diagnostic Mutation DataBase (DmuDB)
• Summary and conclusions

3
Diagnostic Laboratories
ACC
CGH

• Cytogenetics (incl. leukemia)
• 30 sites listed with ACC
• Molecular genetics
• 42 labs list on CMGS website
• 29 regional centers listed with UKGTN
• Molecular cytogenetics (Array CGH)
• 12 labs involved with Microarray CGH
  implementation group

UKGTN
CMGS
4
Data types

• Referral data
• Patient details, phenotype, family details,
  sample details
• Experimental data
• Gel images
• Trace files
• Text files
• Workflow data (worksheets / audit trail)
• Reporting data
• Summarised experimental result
• Mutation call (DNA, protein)
• Interpretation / Clinical significance
• Collateral data
• Polymorphisms?
• Unreported variants?

5
Data storage

• Diagnostics not IT driven
• Internal / local DB
• Varying sophistication
• Referral data ?
• Experimental data ?
• Workflow data ?
• Reporting data ? ?
• Collateral data ? ?
• Very limited sharing of variation data
• Infrastructure
• Time pressures

6
Data usage

• Legal requirement
• Retention and storage of pathological records and
  archives 3rd edition, 2005 recommends at least
  30 years.
• Internal reference
• To aid future diagnosis
• Research
• External reference
• Other diagnostic labs
• Research

7
Data Requirements

• Varies according to disease, reporting policy and
  specific finding.
• Disease specific databases (e.g. BIC)
• Other public databases (e.g. HGMD, DECIPHER)
• Literature searches (usually via DB search)
• Other online tools (e.g. splice site prediction)
• Specific data relating to the proven clinical
  significance of a variant

8
Mutation repository (DmuDB)
9
Issues

• Standardisation
• Mutation nomenclature
• HGVS
• Coping with new versions
• Reference sequences
• Representing phenotypes, Sample and
  interpretation nomenclature
• SNOMED CT (Systematized Nomenclature of Medicine)
• Curation
• Funding
• Control
• Contributors
• Access / publication
• Consent / Confidentiality
• Funding for maintenance
• Lack of diagnostic input into DBs - Inertia

10
BRCA DB example
c.3373_3374delTC
c.3375_3376delTC
11
Nomenclature tool
12
Issues

• Standardisation
• Mutation nomenclature
• HGVS
• Coping with new versions
• Reference sequences
• Representing phenotypes, Sample and
  interpretation nomenclature
• SNOMED CT (Systematized Nomenclature of Medicine)
• Curation
• Funding
• Control
• Contributors
• Access / publication
• Consent / Confidentiality
• Funding for maintenance
• Lack of diagnostic input into DBs - Inertia

13
Summary

• Diagnostic laboratories are producing a large
  amount of high quality data
• This data is largely not shared either within the
  diagnostic community or externally
• NGRL(M) are developing DmuDB as a resource to
  facilitate effective data sharing
• Ownership of data will remain with the user
• Publication of variation data will be encouraged
  by ease of use
• However significant issues remain to be resolved

14
Acknowledgements

• Andrew Devereau
• Ed Burke
• National Genetics Reference Laboratory
  (Manchester)
• St Mary's Hospital
• David Gokhale
• Cheshire Merseyside Genetic Services,
• Liverpool Womens Hospital
• Graham Taylor
• Yorkshire Regional DNA Laboratory
• St James's University Hospital