MolMed Course “ Genetics for Dummies ” Rotterdam, 7 November, 2018 Introduction to Complex Genetics: Concepts and Tools: part B André G Uitterlinden Genetic Laboratory Department of Internal Medicine Department of Epidemiology Department of Clinical Chemistry www.glimdna.org Our website… Professor Trifonius Zonnebloem Professor Cuthbert Calculus Professeur Tryphon Tournesol
DNA polymorphisms Definition: a DNA sequence variation that occurs in the population with a frequency of….. at least ……0.1 % ….0.5%....1%...5%...? Genome Frequency: ~ 1 in 100 base pairs is polymorphic …… probably more! ………………. depends on samples sequenced globally …. 2016: ~500k samples sequenced WES/WGS >150 million polymorphisms in human genome f > 0.1% >50 million polymorphisms f > 1%
Overlap between Rotterdam Study WES coding (refseq) variants* across other publicly available datasets ESP (N=6.503) ExAC (N=60.706) 1000G (N=2.504)) 236 204 972 280 159 829 UK10K (N=3.781) 129 311 5.098 288 151 307 GoNL (N=500) 326 114 61 121 319 5.368 Rotterdam Study WES (60x), n=2,628 samples *Numbers are x1000 Combined databases (N=73.994) Van Rooij, Verkerk, Kraaij, unpublished
Types of DNA polymorphisms (1) How do they look like ? -Single Nucleotide Polymorphisms (SNPs; e.g. A to G) -Insertion/deletion (e.g., AATCGC / -) -Variable Number of Tandem Repeats (VNTRs) *homopolymers: repeat unit = 1 bp (e.g., poly-A) *microsatellites: repeat unit = 2-6 bp (e.g., GT or CA) *minisatellites: repeat unit = >7 bp
Types of DNA polymorphisms (2) What do they do ? exon 3’ UTR 5’ promoter intron Gene structure POSSIBLE FUNCTIONALITY: • 5’promoter: affecting mRNA expression by production • 3’UTR: affecting mRNA expression by stability • intron: affecting mRNA expression, splicing • exons: affecting protein structure/activity
Types of DNA polymorphisms (3) What do they do ? 3’ UTR exon 5’ promoter intron Gene structure “NON - FUNCTIONAL” ??? • exons: synonymous codon changes (e.g., both TCT and TCC encode the aa Serine) • introns • intergenic areas (“gene deserts”) Problem: For all these cases functional SNPs have been identified. >> We have NOT analyzed ALL SNPs in the human genome PROPERLY They could all very well be functional !
SNPs, alleles, genotypes and haplotypes SNP= Single Nucleotide Polymorphism strand Allele A A C G + chromosomes + C G A T Genotype Haplotype Allele A A C G C G A T
Genetic Association Analysis (1) case-control design Test for “association” by counting variants of a (candidate) gene. Compare allele frequencies: A = wild-type allele; B = risk-allele AAAAA BBB AAAAA BBBBB AA 70% 30% 50% 50% DISEASE-group CONTROL-group
Genetic Association Analysis (2) Quantitative Trait analyses Humans are diploid: compare characteristics by their genotype Genotype mean Femoral Neck BMD AA 0.82 ±0.12 0.82 ±0.12 0.82 ±0.12 AB 0.80 ±0.13 0.82 ±0.13 0.79 ±0.13 BB 0.78 ±0.13 0.78 ±0.13 0.79 ±0.13 dose-effect recessive dominant Population (-based sample)
What is a Genome-Wide Association Study? • Method for interrogating millions of common DNA variations across the human genome • Based on classic association study design • GWAS is based on “Linkage Disequilibrium” (LD): >DNA variation is inherited in blocks, so not all variants have to be tested because one/a few will predict others
Candidate Gene Analysis: 1990-2005 Specific Expression Specific Expression GWA GWA Animal Model Animal Model Mendelian Disease Mendelian Disease Identification of of Candidate Candidate Gene Gene Identification Identify D DNA NA polymorphisms polymorphisms Identify Identify haplotypes Identify haplotypes Association analysis Functional analysis Association analysis Functional analysis with disease phenotype phenotype in in in cells cells/serum/etc. /serum/etc. with disease in populations populations Meta- Meta -analysis to quantify analysis to quantify effect effect size size
Candidate gene association analysis “in practice”: Many controversial & ir-reproducible results because of: - Small sample size - Ill-defined choice of polymorphisms - Lack of standardized genotyping - Lack of standardized phenotype data - Publication bias >> How to improve? - Combine study populations (across Europe, globally): meta-analysis - Rationalise choice of polymorphisms: functionality, haplotypes - Standardize genotyping methods: reference DNA plate - Standardize phenotypes across populations: meta-analysis individual level data - Run prospective meta-analyses
IGFI, IGFBP3 GR, 11B-HSD Cortisol IGF/GH TSHR, DIO1, DIO2, DIO3, MCT8 Thyroid Hormone ERα , ERβ, Aromatase, VDR ,DBP LH, LHR, GnRH Vitamin D Estrogen Genetic determinants of osteoporosis ? TGFb/BMP/Wnt- homocysteine signalling Matrix molecules MTHFR, MS, MTRR, CBS, THYMS TGFb, LRP5/6 , Collagen Ia1 , osteocalcin, BMP2, FRZB, AHSG, LOX SOST
“GENOMOS” a large-scale, multi-centre study for prospective meta- analyses of osteoporosis candidate gene variants “Genetic Markers for Osteoporosis” Total number EU FP5 sponsored: 3 mio euro of subjects Jan 2003 – Jan 2007 (early 2006): 26,264 Genes analysed: - ESR1 18,405 women 16 Aberdeen - COLIA1 2 14 7,859 men Aarhus - VDR 13 15 3 - TGFb Cambridge - LRP5&6 5 Amsterdam 12 6,498 fractures Warsaw 1 10 7 Rotterdam* Antwerp 2,380 vertebral fx Graz 11 4 Firenze 9 Barcelona 8 9 = Participant + Epidemiological Cohort Ioannina 7 = Participant *coordinating centre
GENOMOS RESULTS (March 2008) GENE SNPs n Sample n BMD FX (6) (17) FN LS Vert Non-Vert PUBLICATIONS : ESR1 3 18,917 - - 20-30% 10-20% Ioannidis et al., JAMA 2004 COLI 1 20,786 0.15 SD 0.15 SD 10% (Sp1) - Ralston et al., PLoS Med 2006 VDR 5 26,242 - - 10% (Cdx) - Uitterlinden et al., Ann Int Med 2006 TGFb 5 28,924 - - - - Langdahl et al. Bone 2008 LRP5 2 37,760 0.15 SD 0.15 SD 12-26% 6-14% van Meurs et al. JAMA 2008 LRP6 1 37,760 - - - -
EU-FP7 EU FP7 pr proj ojec ect: t: GEFOS EFOS (2008-2012) Number of subjects: GENOMOS: >150,000 of which GWAS: 40,000 = G GENOMOS study dy popul ulat ation on = i idem + GWAS www.gefos.org = i idem, m, under er negot otiati ation on / i in devel elopm opment ent
GEFOS HYPOTHESIS-FREE GWAS: AS SAMPLE SIZE INCREASES, GENOME-WIDE SIGNIFICANT SIGNALS BECOME GRADUALLY EVIDENT LUMBAR SPINE BMD 5 x 10 -8 • Rotterdam Study • ERF Study N=5,000 • Twins UK • deCODE Genetics • Framingham Study Rivadeneira et al., Nat Genet., 2009
LUMBAR SPINE BMD LRP5 5 x 10 -8 • Rotterdam Study • ERF Study N=6,200 • Twins UK • deCODE Genetics • Framingham Study Rivadeneira et al., Nat Genet., 2009
LUMBAR SPINE BMD LRP5 5 x 10 -8 • Rotterdam Study • ERF Study N=8,500 • Twins UK • deCODE Genetics • Framingham Study Rivadeneira et al., Nat Genet., 2009
LUMBAR SPINE BMD RANK-L C6ôrf10 OPG 1p36 LRP5 5 x 10 -8 MHC • Rotterdam Study • ERF Study N=15,000 • Twins UK • deCODE Genetics • Framingham Study Rivadeneira et al., Nat Genet., 2009
LUMBAR SPINE BMD RANK-L C6ôrf10 OPG 1p36 SP7 LRP5 5 x 10 -8 • Rotterdam Study • ERF Study N=19,125 • Twins UK • deCODE Genetics • Framingham Study Rivadeneira et al., Nat Genet., 2009
GEFOS collaboration has generated many GWAS discoveries for BMD… Slide by Fernando Rivadeneira
Even bigger sample size yields many more discoveries for both common and less frequent variants……. >300 ~240 2016 UK BIOBANK n= 160K 1000GP (Unpublished data)
UK Biobank: Largest BMD* GWAS so far… in 426,824 White-British participants – 1,103 conditionally independent SNPs from 515 loci (301 novel) – 2x the previous study! – 20% of the trait variance explained – 1.5x increase! *estimated BMD from heel QUS Slide by John Morris; Morris J, Kemp J et al. Nature Genetics (acepted)
Recommend
More recommend
