Transcript Julian Sampson.ppsx

Recent Advances
in Genomic Science
Julian Sampson
Institute of Medical Genetics, Cardiff
“the human genome sequence offers a
unique opportunity to understand
genetic factors in health and disease,
and to apply this rapidly to prevention,
diagnosis and treatment”
Francis Collins, Director NHGRI (and now of NIH)
to US House of Representatives, May 2003
April 24th 2003
“the human genome sequence offers a
unique opportunity to understand
genetic factors in health and disease,
and to apply this rapidly to prevention,
diagnosis and treatment”
Francis Collins, Director NHGRI
to US House of Representatives, May 2003
April 24th 2003
How Far Have We Got ?
• Changing technologies
• Set the scene for discussion of
application: linking genomic
variation and disease (to inform
diagnosis, prevention, treatment)
Changing Technology for Testing the Genome:
Resolution, Scale, Speed and Cost
Karyotype
• 5-10Mb (≈ 107bp)
• Several weeks
• Banding from 1960s
Changing Technology for Testing the Genome:
Resolution, Scale, Speed and Cost – 1990s
• Resolution depends on
probe density
• Days
Whole genome
or targeted aCGH
Karyotype
• 5-10Mb (≈ 107bp)
• Several weeks
•
•
•
PCR amplicon
1bp Resolution
Extremely accurate
Automated Sanger Sequencing
Testing the Genome with DNA Arrays
Reference
genomic DNA
Test sample genomic
DNA
Cy3
Arrayed DNA sequences
(oligos)
Cy5
Image
individual spots
(n = 2M in DDD)
• Cy3/Cy5 ratio calculated for each arrayed sequence
• Identifies deletions or duplications in genome
(copy number variants or CNVs)
aCGH: Child with Seizures, Microcephaly and
Developmental Delay
Chromosome 14
14q12
14q12
14q12
≥2.3kb
Deletion
aCGH: A 2.3kb deletion of FOXG1 at 14q12
WALES LABORATORY CALL HISTORY (n=2)
DE NOVO DELETED REGION IN THE PATIENT
FOXG1
GENES
• DECIPHER database – over 30 overlapping deletions “FOXG1 syndrome”
• (Decipher has approx. 20,000 entries from >30 countries)
- severe developmental delay, brain malformation, seizures, microcephaly
• Current aCGH detects a genetic cause in
5-20% of patients with developmental disorders
•But: CNVs that Non-Pathogenic or that are associated with predisposition/variability
Changing Technology for Testing the Genome:
Resolution, Scale, Speed and Cost
• <1 kb (≈ 102 bp)
• Several days
Whole genome
or targeted aCGH
Karyotype
• 5-10Mb (≈ 107bp)
• Several weeks
• Sanger sequencing
• 1st Genome:
• $3Billion, 13 Years
• Follow up genomes $100M
Changing Technology for Testing the Genome:
Resolution, Scale, Speed and Cost
Whole genome
or targeted aCGH
Karyotype
• 5-10Mb (≈ 107bp)
• Several weeks
NGS: genes,
exomes, genomes
• Genome at 1bp
resolution
• Days (hours)
• “$1000”
• IT / Bioinformatics
Falling Costs of Sequencing a Human Genome (log scale)
Characterising Genomic Variation by NGS
Sequence:
• gene, gene panels (e.g. retina 108, epilepsy 31),
exomes, genomes
Identify and Characterise:
• SNPs, insertions, deletions
• Translocations, inversions
• CNVs and Aneuploidies
i.e. NGS will do virtually everything other
technologies can do, and at 1bp resolution
NGS technologies Lack Specificity
• Mis-calling of bases 0.1-1%
• Need to distinguish true variants from artifacts
• “Read depth” important, but varies across
exome/genome
• → filtering algorithms (“variant calling
pipelines”)
• Bioinformatics expertise is in research centres,
not in the NHS (more joint working needed)
Genome Variation: Constitutional & Somatic
Constitutional
(germline)
variation
Cancer Genomes
Genomic Instability and
somatic variation
Genome Variation: Constitutional & Somatic
Constitutional
(germline)
variation
Cancer Genomes
Genomic Instability and
somatic variation
The prevalence of somatic mutations
across human cancer types.
LB Alexandrov et al. Nature (2013)
Cancer Genomes:
• Genomic instability creates heterogeneous cell
populations (many differently evolving clones)
• Critical mutations (e.g. for drug resistance)
may be present in sub-set of cells
• Distinguishing low level mutations from miscalled bases is bioinformatically challenging –
variants require validation
Genome Variation: Neutral & Disease-Associated
Constitutional
(germline)
variation
Cancer Genomes and
somatic variation
Functional and Polymorphic Variation
“Passenger” and “Driver” Mutations
Linking Genomic Data and Disease
• Distinguish disease-related from neutral variation
• 250 – 300 loss of function mutations per genome in
annotated genes
• Missense variants
• CNVs
• Databases of genomic variation and phenotype data
• Statistical, in silico, in vitro and in vivo approaches
Cataloguing Genomic Variation
and relating this to disease
•
•
•
•
•
•
•
•
Decipher
Human Gene Mutation Database (HGMD)
Human Genome Variation Project (HGV)
1000 Genomes
UK10K
100,000 Genomes (UK)
Cosmic
The Cancer Genome Atlas (TCGA)
Cataloguing Genomic Variation
and relating this to disease
•
•
•
•
•
•
•
•
Decipher
Human Gene Mutation Database (HGMD)
Human Genome Variation Project (HGV)
1000 Genomes
UK10K
100,000 Genomes (UK)
Cosmic
The Cancer Genome Atlas (TCGA)
Sequencing for unknown
disease-causing variants: Trios
de novo mutations
(e.g. in Exome or
Genome)
e.g. Mendelian and developmental disorders (DDD
project), autism, schizophrenia
Inherited “Single Gene” (Mendelian)
Disorders
• >20,000 genes in the human genome
• > 7,000 Mendelian disorders
• 1 in 17 people have a “rare disease” – i.e. one
that affects < 1 in 2000 of the population
• Individually rare, cumulatively frequent
• Many genes identified, pathophysiology
becoming understood, targeted treatments
emerging..
Complex Disorder
Multiple Causes
Shared
Phenotype
Mendelian Disorder
Single
Gene
Multiple phenotypic effects
Mendelian Disorder
Complex Disorder
Stratified Medicine
Shared
Phenotype
Single
Gene
Targeted Treatment
Multiple Causes
Multiple phenotypic effects
Many Genomes are Relevant to Human Health
• Human Genome
• Model organisms, pathogens and vectors
Many “-Omic” Applications
of Next Generation Sequencing
• Genetics and Genomics
• Epigenomics (e.g. “methylome”) -probing a
mechanism for regulating and adapting the genome
• Transcriptomics – probing differential genome usage
(time, place, environment)
Cho et al. Nature 2012
Costello et al. Nat Biotech 2009
Summary
• Cost of genomic analysis in healthcare is now
as affordable as many other technologies
• Benefits in mendelian / chromosomal
disorders and stratified medicine already
translating from research to the clinic
• Diagnostics well developed
• Targeted therapy and prevention based upon
genomic understanding gaining pace