genome-wide association studies: hunting for genes in the new millennium

Genome-Wide Association Studies: Hunting for

Genes in the New Millennium

National Human Genome Research

Institute

National Institutes of

Health

U.S. Department of Health and

Human Services

U.S. Department of Health and Human Services

National Institutes of HealthNational Human Genome Research

InstituteTeri A. Manolio, M.D., Ph.D.Director, Office of Population GenomicsSenior Advisor to the Director, NHGRI,

for Population GenomicsNovember 20, 2008

We Live in Interesting Times…

“‘May he live in interesting times.’ Like it or not we live in interesting times.”

--Robert Kennedy, June 7, 1966

May you come to the attention of those in authority.

May you find what you are looking for.

Wikipedia, accessed 11Sep07

200520062007 first quarter2007 second quarter2007 third quarter2007 fourth quarter2008 first quarter 2008 second quarter 2008 third quarter

Manolio, Brooks, Collins, J. Clin. Invest., May 2008

Pennisi E, Science 2007; 318:1842-43.

2007: The Year of GWA Studies

Diseases and Traits with Published GWA Studies (n = 76, 11/17/08)

• Macular Degeneration• Exfoliation Glaucoma

• Lung Cancer• Prostate Cancer• Breast Cancer• Colorectal Cancer• Bladder Cancer• Neuroblastoma• Melanoma• TP53 Cancer

Predispos’n• Chr. Lymph. Leukemia

• Inflamm. Bowel Disease• Celiac Disease• Gallstones• Irritable Bowel

Syndrome

• QT Prolongation • Coronary Disease• Coronary Spasm • Atrial

Fibrillation/Flutter• Stroke• Subarachnoid

Hemorrhage• Intracranial Aneurysm • Hypertension• Hypt. Diuretic Response• Peripheral Artery

Disease

• Syst. Lupus Erythematosus

• Sarcoidosis• Pulmonary Fibrosis• Psoriasis • HIV Viral Setpoint• Childhood Asthma

• Type 1 Diabetes • Type 2 Diabetes• Diabetic Nephropathy • End-St. Renal Disease• Obesity, BMI, Waist,

IR• Height• Osteoporosis• Osteoarthritis• Male Pattern Baldness

• F-Cell Distribution• Fetal Hgb Levels• C-Reactive Protein• ICAM-1• Total IgE Levels• Uric Acid Levels, Gout• Protein Levels• Vitamin B12 Levels• Recombination Rate• Pigmentation

• Lipids and Lipoproteins• Warfarin Dosing• Ximelegatran Adv.

• Parkinson Disease• Amyotrophic Lat.

Sclerosis• Multiple Sclerosis• MS Interferon-β

Response • Prog. Supranuclear

Palsy• Alzheimer’s Disease in

ε4+• Cognitive Ability• Memory• Hearing• Restless Legs Syndrome • Nicotine Dependence• Methamphetamine

Depend.• Neuroticism• Schizophrenia• Sz. Iloperidone

Response• Bipolar Disorder• Family Chaos• Narcolepsy• Attention Deficit

Hyperactivity• Personality Traits

• Rheumatoid Arthritis• RA Anti-TNF Response

Hunter DJ and Kraft P, N Engl J Med 2007; 357:436-439.

“There have been few, if any, similar bursts of discovery in the history of medical research…”

What is a Genome-Wide Association Study?

• Method for interrogating all 10 million variable points across human genome

• Variation inherited in groups, or blocks, so not all 10 million points have to be tested

• Blocks are shorter (so need to test more points) the less closely people are related

• Technology now allows studies in unrelated persons, assuming 5,000 – 10,000 base pair lengths in common (300,000 – 1,000,000 markers)

DNA on Chromosome 7GAAATAATTAATGTTTTCCTTCCTTCTCCTATTTTGTCCTTTACTTCAATTTATTTATTTATTATTAATATTATTATTTTTTG

AGACGGAGTTTC/ACTCTTGTTGCCAACCTGGAGTGCAGTGGCGTGATCTCAGCTCACTGCACACTCCGCTTTCCTGGTTTCAAGCGATTCTCCTGCCTCAGCCTCCTGAGTAGCTGGGACTACAGTCACACACCACCACGCCCGGCTAATTTTTGTATTTTTAGTAGAGTTGGGGTTTCACCATGTTGGCCAGACTGGTCTCGAACTCCTGACCTTGTGATCCGCCAGCCTCTGCCTCCCAAAGAGCTGGGATTACAGGCGTGAGCCACCGCGCTCGGCCCTTTGCATCAATTTCTACAGCTTGTTTTCTTTGCCTGGACTTT

ACAAGTCTTACCTTGTTCTGCC/TTCAGATATTTGTGTGGTCTCATTCTGGTGTGCCAGTAGCTAAAAATCCATGATTTGCTCTCATCCCACTCCTGTTGTT

CATCTCCTCTTATCTGGGGTCACA/CTATCTCTTCGTGATTGCATTCTGATCCCCAGTACTTAGCATGTGCGTAACAACTCTGCCTCTGCTTTCCCAGGCTGTTGATGGGGTGCTGTTCATGCCTCAGAAAAATGCATTGTAAGTTAAATTATTAAAGATTTTAAATATAGGAAAAAAGTAAGCAAACATAAGGAACAAAAAGGAAAGAACATGTATTCTAATCCATTATTTATTATACAATTAAGAAATTTGGAAACTTTAGATTACACTGCTTTTAGAGATGGAGATGTAGTAAGTCTTTTACTCTTTACAAAATACATGTGTTAGCAATTTTGGGAA

GAATAGTAACTCACCCGAACAGTG/TAATGTGAATATGTCACTTACTAGAGGAAAGAAGGCACTTGAAAAACATCTCTAAACCGTATAAAAACAATTACATCATAATGATGAAAACCCAAGGAATTTTTTTAGAAAACATTACCAGGGCTAATAACAAAGTAGAGCCACATGTCATTTATCTTCCCTTTGTGTCTGTGTGAGAATTCTAGAGTTATATTTGTACATAGCATGGAAAAATGAGAGGCTAGTTTATCAACTAGTTCATTTTTAAAAGTCTAACACATCCTAGGTATAGGTGAACTGTCCTCCTGCCAATGTATTGCACATTTGTGCCCAGATCCAGCATAGGGTATGTTTGCCATTTACAAACGTTTATGTCTTAAGAGAGGAAATATGAAGAGCAAAACAGTGCATGC

TGGAGAGAGAAAGCTGATACAAATATAAAT/GAAACAATAATTGGAAAAATTGAGAAACTACTCATTTTCTAAATTACTCATGTATTTTCCTAGAATTTAAGTCTTTTAATTTTTGATAAATCCCAATGTGAGACAAGATAAGTATTAGTGATGGTATGAGTAATTAATATCTGTTATATAATATTCATTTTCATAGTGGAAGAAATAAAATAAAGGTTGTGATGATTGTTGATTATTTTTTCTAGAGGGGTTGTCAGGGAAAGAAATTGCTTTTT

SNPs 1 / 300 bases

Christensen and Murray, N Engl J Med 2007; 356:1094-97.

Mapping the Relationships Among SNPs

Chromosome 9p21 Region Associated with MI

Samani N et al, N Engl J Med 2007; 357:443-453.

BostonProvi-dence

New York

Phila-delphi

Balti-more

Providence 59

New York 210 152Philadelphia 320 237 86

Baltimore 430 325 173 87Washington 450 358 206 120 34

Distances Among East Coast Cities

BostonProvi-dence

New York

Phila-delphi

Balti-more

Providence 59

< 100 101-200

201-300

301-400

BostonProvi-dence

New York

Phila-delphi

Balti-more

Providence 59

< 100 101-200

201-300

301-400

Boston Provi-

dence New

York Phila-

delphia Balti-

more Wash-

ington

Boston

Provi-

New York

Phila-delph

Balti-more

ington

Christensen and Murray, N Engl J Med 2007; 356:1094-97.

Mapping the Relationships Among SNPs

One Tag SNP May Serve as Proxy for Many

CAGATCGCTGGATGAATCGCATCTGTAAGCAT

CGGATTGCTGCATGGATCGCATCTGTAAGCAC

CAGATCGCTGGATGAATCCCATCAGTACGCAT

CGGATTGCTGCATGGATCCCATCAGTACGCAT

CGGATTGCTGCATGGATCCCATCAGTACGCAC

SNP2↓

SNP3↓

SNP4↓

SNP5↓

SNP6↓

SNP1↓

Block 1 Block 2

SNP7↓

SNP8↓

SNP2↓

SNP3↓

SNP4↓

SNP5↓

SNP6↓

SNP1↓

Block 1 Block 2

SNP7↓

SNP8↓

SNP3↓

SNP5↓

SNP6↓

Block 1 Block 2

SNP7↓

SNP8↓

SNP3↓

SNP6↓

Block 1 Block 2

SNP8↓

GTT 35%

CTC 30%

GTT 10%

GAT 8%

CAT 7%

CAC 6%

other haplotypes 4%

Block 1 Block 2 FrequencySingleton

www.hapmap.org

Nature 2005; 437:1299-320.Nature 2007; 449:851-61.

A HapMap for More Efficient Association Studies: Goals

• Use just the density of SNPs needed to find associations between SNPs and diseases

• Do not miss chromosomal regions with disease association

• Produce a tool to assist in finding genes affecting health and disease

• Use more SNPs for complete genome coverage of populations of recent African ancestry populations due to shorter LD

Progress in Genotyping Technology

1 10 102 103 104 105 106

Nb of SNPs

ABITaqMan

ABISNPlex

IlluminaGolden

IlluminaInfinium/

Sentrix Affymetrix

100K/500K

Perlegen

Affymetrix

MegAllele

2001 2005

Affymetrix

Courtesy S. Chanock, NCI

Jul-05 Oct-05 Jan-06 Apr-06 Jul-06 Oct-06

Affymetrix 500K

Illumina 317K

Illumina 550K

Illumina 650Y

Continued Progress in Genotyping Technology

Courtesy S. Gabriel, Broad/MIT

July 2005 Oct 2006

Association of Alleles and Genotypes of rs1333049 with Myocardial

Infarction C

N (%)G

N (%)2

(1df)P-value

Cases2,132 (55.4)

1,716 (44.6)

55.11.2 x 10-

13Controls

2,783 (47.4)

3,089 (52.6)

Allelic Odds Ratio = 1.38

Association of Alleles and Genotypes of rs1333049 with Myocardial

Infarction C

N (%)G

N (%)2

(1df)P-value

Cases2,132 (55.4)

1,716 (44.6)

55.11.2 x 10-

13Controls

2,783 (47.4)

3,089 (52.6)

Allelic Odds Ratio = 1.38CC

N (%)CG

N (%)GG

N (%)2

(2df) P-value

Cases586

(30.5) 960 (49.9)

378 (19.6)59.7

1.1 x 10-

14Controls

676 (23.0)

1,431 (48.7)

829 (28.2)

Heterozygote Odds Ratio = 1.47

Homozygote Odds Ratio = 1.90

Klein et al, Science 2005; 308:385-389.

P Values of GWA Scan for Age-Related Macular Degeneration

Bierut LJ et al, Hum Molec Genet 2007; 16:24-35.

Nicotine Dependence among Smokers

http://www.broad.mit.edu/diabetes/scandinavs/type2.html

Genome-Wide Scan for Type 2 Diabetes in a Scandinavian Cohort

Libioulle C et al, PLoS Genet; 2007 Apr 20;3(4):e58.

Genome-Wide Scan for Crohn Disease in Belgian Cases and

Controls

Sladek R et al, Nature 2007; 445, 881-885.

Genome-Wide Scan for Type 2 Diabetes in French Case-Control

WTCCC, N ature 2007; 447:661-678.

Wellcome Trust Genome-Wide Association Study of Seven

Common Diseases

Hunter DJ et al, Nat Genet 2007; 39:870-874.

Genome-Wide Scan for Breast Cancer in Postmenopausal Women

-Log10 P Values for SNP Associations with Myocardial

Infarction

Samani N et al., N Engl J Med 2007; 357:443-53.

Association Signal for Coronary Artery Disease on Chromosome 9

Samani N et al., N Engl J Med 2007; 357:443-53.

Region of Chromosome 1 Showing Strong Association with Inflammatory

Bowel Disease

Duerr R et al., Science 2006; 314:1461-63.

Unique Aspects of GWA Studies

• Permit examination of inherited genetic variability at unprecedented level of resolution

• Permit "agnostic" genome-wide evaluation • Once genome measured, can be related to any

trait• Most robust associations in GWA studies have not

been with genes previously suspected of association with the disease

• Some associations in regions not even known to harbor genes

“The chief strength of the new approach also contains its chief problem: with more than 500,000 comparisons per study, the potential for false positive results is unprecedented.”

Hunter DJ and Kraft P, N Engl J Med 2007; 357:436-439.

Larson, G. The Complete Far Side. 2003.

Number of New, Significant Gene-Disease Associations by Year, 1984 -

Hirschhorn J et al, Genet Med 2002; 4:45-61.

Of 600 Gene-Disease Associations, Only 6 Significant in > 75% of

Identified Studies

Disease/Trait GenePolymorphism

Frequency

DVT F5 Arg506Gln 0.015

Graves’ Disease

CTLA4 Thr17Ala 0.62

Type 1 DM INS 5’ VNTR 0.67

HIV/AIDS CCR5 32 bp Ins/Del 0.05-0.07

Alzheimer’s APOE Epsilon 2/3/4 0.16-0.24

Creutzfeldt-Jakob Disease

PRNP Met129Val 0.37

Hirschhorn J et al., Genet Med 2002; 4:45-61.

POLYMORPHISM PRESENT ABSENT SUMMARY

ACE I/D 13 with D; 1 with I 18 favors none

APOE 8 with ε4, 2 with ε2 9 equivocal

AGT M235T 0 8 none

AGTR1 A1166C 0 7 none

MTHFR 7 with T, 1 with non-T 8 equivocal

PON1 Q192R 3 with R 10 none

PON1 L55M 5 with L (subgroups) 1 weak

NOS3 G894T 1 with T 4 none

MMP3 -1516 5A/6A

4 with 6A0

association

IL-6 G-174C 1 with G 3 none

Reports For and Against Associations of Variants with

Carotid Atherosclerosis

Manolio et al., ATVB 2004; 24:1567-77.

May 1999

J. Hirschhorn and D. Altshuler J Clin Endo Metab 2002

Am J Hum Genet July 2004

PLoS Biol Sept 2005

Nat Genet July 2006

Chanock S, Manolio T, et al., Nature 2007; 447:655-60.

Replication, Replication, ReplicationInitial study: Sufficient description to permit

replication• Sources of cases and controls• Participation rates and flow chart of selection• Methods for assessing affected status• Standard “Table 1” including rates of missing data• Assessment of population heterogeneity• Genotyping methods and QC metricsReplication study: • Similar population, similar phenotype• Same genetic model, same SNP, same direction• Adequately powered to detect postulated effect

Chanock S, Manolio T, et al., Nature 2007; 447:655-60.

Replication Study #1 3,000 cases / 3,000 controls

Replication Study #22,400 cases / 2,400 controls

Replication Study #3 2,500 cases / 2,500 controls

Initial Study1,150 cases / 1,150 controls

~24,000 SNPs

~1,500 SNPs

200+ New ht-SNPs

>500,000 Tag SNPs

25-50 Loci

Replication Strategy for Prostate Cancer Study in CGEMS

Hoover R, Epidemiology 2007; 18:13-17.

Replication Strategy in Easton Breast Cancer Study

Stage Cases Controls SNPs

1 408 400 266,722

Easton et al, Nature 2007; 447:1087-93.

1 408 400 266,722

2 3,990 3,916 13,023

1 408 400 266,722

2 3,990 3,916 13,023

3 23,734 23,639 31

• ABCFS• BCST• COPS • GENICA• HBCS• HBCP

1 408 400 266,722

2 3,990 3,916 13,023

3 23,734 23,639 31

Final 6

• MEC-W• MEC-J• NHS• PBCS• RBCS• SASBAC

• SEARCH2• SEARCH3• SBCP• SBCS• CNIOBCS• USRT

• TBCS• KConFab/AOCS• KBCP• LUMCBCS• MCBCS• MCCS

Larson, G. The Complete Far Side. 2003.

Replication Strategy in CGEMS Prostate Cancer Study

1 1,172 1,157 527,869

Thomas et al, Nat Genet 2008; 40:310-15.

1 1,172 1,157 527,869

2 3,941 3,964 26,958*

1 1,172 1,157 527,869

2 3,941 3,964 26,958*

* Selected for p < 0.068

1 1,172 1,157 527,869

2 3,941 3,964 26,958*

SNP GeneStage 1+2

P-value

rs4962416 MSMB 7 x 10-13

rs10896449 11q13 2 x 10-9

rs10993994 CTBP2 2 x 10-7

rs10486567 JAZF1 2 x 10-6

1 1,172 1,157 527,869

2 3,941 3,964 26,958*

SNP GeneStage 1+2

P-value

Initial Rank

rs4962416 MSMB 7 x 10-13 24,223

rs10896449 11q13 2 x 10-9 2,439

rs10993994 CTBP2 2 x 10-7 319

rs10486567 JAZF1 2 x 10-6 24,407Thomas et al, Nat Genet 2008; 40:310-15.

1 1,172 1,157 527,869

2 3,941 3,964 26,958*

SNP GeneStage 1+2

P-valueInitial Rank

Initial P-value

rs4962416 MSMB 7 x 10-13 24,223 0.042

rs10896449 11q13 2 x 10-9 2,439 0.004

rs10993994 CTBP2 2 x 10-7 319 4 x 10-4

rs10486567 JAZF1 2 x 10-6 24,407 0.042

2005 2006 2007 2008

Published GWA Reports, 3/2005 - 9/2008

Calendar Quarter

NHGRI Catalog of GWA Studies: http://www.genome.gov/gwastudies/

NHGRI GWA Catalog - Objectives

• Identify and track all GWA publications attempting to assay > 100,000 SNPs

• Extract key information regarding associations

• Provide widely as scientific resource, including downloadable datafile

• Seek commonalities across associations genome-wide rather than disease by disease

• Describe approach clearly so others can replicate or expand upon it

• Maintain consistency in approach• Adapt to evolving technologies: CNVs?

NHGRI GWA Catalog - Methods

• Survey NIH “e-clips” daily, PubMed weekly• Identify all GWA publications attempting to

assay > 100,000 SNPs• Describe top 5 novel associations significant

at p < 10-6 • Expand to all associations < 10-6 • Extract information on:

- Disease/trait - Rs number/risk allele - Sample size - Risk allele frequency- Genomic region - P-value, OR [95%

CI]- Reported genes - Platform, #SNPs

Reports Included in this Analysis

• 180 published papers through 9/18/2008

• 34 did not report SNP

• 1 reported haplotypes without specific SNPs

• 145 reports

– 782 unique (“index”) SNPs

– 3,841 unique perfect LD SNPs (“linked”)

– 4,623 index + linked

• 83 index SNPs reported 2-7 times

• 10 multiple reports in “unrelated” traits

0 10 20 30 40 50 60

Intergenic

3' (0.5kb)

5' (2kb)

3' UTR

5' UTR

Intronic

Synonymous

Missense

Functional Classification of 782 Index SNPs Associated with

Complex Traits37

0 10 20 30 40 50 60 Percent

0 10 20 30 40 50 60

Intergenic

3' (0.5kb)

5' (2kb)

3' UTR

5' UTR

Intronic

Synonymous

Missense

Functional Classification of 782 Index SNPs and 4,623 Index +

Linked SNPs

Index SNPs Index + Linked SNPs

0 10 20 30 40 50 60 Percent

Odds Ratios of Discrete Associations

1.2 1.4 1.6 1.8 2.0 2.2 2.4 3.04.05.06.0 9.013.020.030.0

Odds Ratio (upper inclusive bound)

ssocia

3 4 5 6 9 13 20 30

Median = 1.28

Percent of Variance in Disease Risk Explained by 32 Established CD Risk Loci

Barrett et al., Nat Genet 2008 Jun 29.

Power to detect risk loci

Odds Ratios of Discrete Associations

1.2 1.4 1.6 1.8 2.0 2.2 2.4 3.04.05.06.0 9.013.020.030.0

Odds Ratio (upper inclusive bound)

ssocia

3 4 5 6 9 13 20 30

Reported Risk Allele Frequencies by Odds Ratios for

Discrete Traits

0.0 0.2 0.4 0.6 0.8 1.0

Risk Allele Frequency (%)

Discrete Traits

0.0 0.2 0.4 0.6 0.8 1.0

Discrete Traits

0.0 0.2 0.4 0.6 0.8 1.0

Discrete Traits

0.0 0.2 0.4 0.6 0.8 1.0

Thorlieifsson Exfoliation Glaucoma

SarasqueteOsteonecrosis

HakonarsonType 1 DM

van HeelCeliac Disease

WTCCCType 1 DM

Author Trait # Ca/Co RAF OR P-value

Thorliefsson Exfol’n glaucoma 75/14,747 0.85 20.10 3 x 10-21

Sarasquete Osteonecrosis 21/64 0.12 12.75 1 x 10-6

HakonarsonM Type 1 diabetes 561/1,143 0.13 8.30 1 x 10-16

van HeelM Celiac disease 991/1,489 0.14 7.04 1 x 10-19

Matarin* Stroke 259/269 NR 5.62 6 x 10-6

WTCCCM Type 1 diabetes 1,963/2,938 0.39 5.49 5 x 10-134

Behrens Juvenile arthritis 130/1,952 NR 5.37 2 x 10-10

Fung Parkinson’s dis. 267/270 NR 5.00 7 x 10-6

Klein Macular degen. 96/50 0.70 4.60 4 x 10-8

SEARCH Statin myopathy 85/90 0.13 4.50 2 x 10-9

*2 other SNPs in this study also associated, OR > 2.0.MSNPs in MHC region.

Characteristics of SNPs Associated with

Odds Ratios > 4.5

FST Values: Index SNPs and HapMap SNPs

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80

HapMap CEU Index SNPsFST Values

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80

Median = 0.069

Phenotype Relationships of SNPs with Highest FST Values

IR OB Height Cancer

Top 5% Fst values (0.49) Top 1% Fst values (0.69)

Immune Pigment Obesity Neuro. Height BMD Cancer Related Traits Related

Lessons Learned from Initial GWA Studies

Signals in Previously Unsuspected Genes

Macular Degeneration

Coronary Disease CDKN2A/2BChildhood Asthma ORMDL3Type II Diabetes CDKAL1Crohn’s Disease ATG16L1

Signals in Gene “Deserts”Prostate Cancer 8q24

Crohn’s Disease5p13.1,

1q31.2, 10p21

Signals in Common

Coronary Disease CDKN2A/2BDiabetes, Melanoma

Childhood Asthma ORMDL3 Crohn’s DiseaseType II Diabetes CDKAL1 Prostate CancerCrohn’s Disease ATG16L1

Signals in Gene “Deserts”Prostate Cancer 8q24

1q31.2, 10p21

Signals in Common

Signals in Gene “Deserts”Signals in Common

Prostate Cancer 8q24Breast, Colorectal Cancer; Crohn’s

1q31.2, 10p21

Signals in Common

1q31.2, 10p21Signals in Common

Multiple Sclerosis IL7R Type 1 DiabetesSarcoidosis C10orf67 Celiac DiseaseRA, T1DM PTPN2, PTPN22 Crohn’s

Signals in Common

1q31.2, 10p21Signals in Common

Multiple Sclerosis IL7R Type 1 DiabetesSarcoidosis C10orf67 Celiac DiseaseRA, T1DM PTPN2, PTPN22 Crohn’s

Study Crohn’s Disease!

Barrett et al., Nat Genet 2008 Jun 29.

• Nearly half of GWA-identified SNPs are intergenic• Only 8.4% of index SNPs are in coding regions, 5’

or 3’ UTR, or miRTS• Potential selection bias in genotyped SNPs for

excess of missense variants • Most associated odds ratios are < 1.5• Risk allele frequencies do not appear skewed

toward rare alleles or large FST values

• Highly-differentiated SNPs enriched for immune-related, pigmentation, and obesity traits

• Examination of loci at extremes of these characteristics may yield interesting insights

Conclusions

“The more we find, the more we see, the more we come to learn.

The more that we explore, the more we shall return.”

Sir Tim Rice, Aida, 2000

genome-wide association studies: hunting for genes in the new millennium

Documents

the education millennium development goal education...

snark hunting using millennium for acrl annual statistics...

millennium corporate...

genome-wide association studies: hunting for genes in the...

instructions millennium by hunting solutions user...

japanese whale hunting and canada harp seal hunting

millennium serials presentation stefanie wittenbach february...

hunting human disease genes: lessons from the past ... ·...

genes function: genes in action

hunting and hunting season comes again magazine

getting started hunting! · getting started hunting!...

millennium pharmaceuticals, inc. product pipeline …sample...

ohio controlled hunting opportunities...

millennium coal pty ltd millennium expansion …...3 1...

the millennium primary years - tmpy the millennium … ·...

millennium circulation development update millennium silver

millennium lighting, inc. millennium lighting 922 dailey

millennium project the millennium project: a plan for

millennium one step foamable adhesive · millennium one...

[millennium development goals 1 and 3] by [author name...