Background

a lightning tour of genetics

The human genome

• 23 pairs of chromosomes
• 3 billion bases
• 22000 genes
• 1/1350 bases different btw individuals

SNPs

300 million total SNPs (1 every 300 bases)!

SNPs

at Broad Institute

SNPs - traits

height

AA

AG

GG

height

AA

AG

GG

SNPs - disease

AA

AG

GG

Dx

No dx

A

G

Dx

No dx

OR = 11.1

Odds Ratio (OR):

20/6

6/20

SNPs - disease

AA

AG

GG

Dx

No dx

OR = 11.1

Prevalence of disease: 2%

Prevalence of disease w/ SNP: 2% * 11.1 = 22.2%

GWAS Era

GWAS everything

GWAS everything

500K people, 4203 traits

1mil people

5mil people

SNPs (outliers)

• Breast cancer:
  • OR = 3.6-6.9

SNPs (outliers)

• Alzheimer's Disease:

ApoE4 OR:

one copy = 3

two copies = 14.5

 

10-15% of population

from Scientific American

SNPs (outliers)

• Warfarin processing:

Results (typical)

• Schizophrenia is ~80% genetic
• 37K cases, 113K controls

Results (typical)

Results (typical)

Results (typical)

Rate of schizophrenia: 0.25% - 0.64%

rate * 1.32 = 0.33% - 0.84%

Complex traits

Traits

albinism

wet/dry earwax

lactase persistence

 

eye color

hair color

 

height

weight

intelligence

schizophrenia

cancer

Height as a trait

• 60-80% genetic
  • 50% from SNPs
• Top SNPs found through GWAS only explain <1% of height variability, individually

Height as a trait

• 60-80% genetic
  • 50% from SNPs
• Top SNPs found through GWAS only explain <1% of height variability, individually

Idea:

Look at the effects of multiple SNPs together

Polygenic Risk Scores

To calculate polygenic risk scores, add up ORs for every SNP you have

 

Assuming I have one of each SNP above, my PRS is 1.07 + 1.07 + 1.07 + .933 + 1.07 = 5.213

Polygenic Risk Scores

from Whiffin et al

Polygenic Risk Scores

from Whiffin et al

Risk?

80/100 => 80% chance of case-status

80 cases

20 ctls

Height as a trait

• 60-80% genetic
  • 50% from SNPs
• Top SNPs found through GWAS only explain <1% of height variability, individually

from Wood et al

• Using 10,000 SNPs, a polygenic risk score model explained 30% of height variability

Height as a trait

from Sexton et al

Weight as a trait

Polygenic Risk Scores

Potential in Medicine

from Khera et al

Coronary artery disease

Potential in Medicine

from Khera et al

Afib

T2D

B.C.

IBD

PRS for you

PRSs for DTC genetics users

PRSs for DTC genetics users

PRSs for DTC genetics users

How to interpret

Intelligence:

80% genetic,

~5% of variation explained

Eye color:

100% genetic,

~100% of variation explained

Height:

80% genetic,

~20% of variation explained

"Variation explained"

Polygenic Risk Score

Polygenic Risk Score

Brown eyes

Blue eyes

IQ

PRS for personality and psych disorders

History of personality

Sir Francis Galton founded the lexical hypothesis of personality (1884):

 

Important personality characteristics will eventually become part of language

History of personality

• Gordon Allport and Henry Odbert collected 4,504 adjectives  relating to personality (1936)
• Raymond Cattell eliminates synonyms creating 171 terms (1940), found 16 factors
• Reduced to 5 factors by Ernest Tupes and Raymond Christal (1960s)
• In 1980s, Lewis Goldberg and others championed the five-factor model

The Big Five

5-10% genetic

psychRisk

http://leobrueg.shinyapps.io/psychRisk

psychRisk

psychRisk

Additional Resources

DNA.land: dna.land

Impute.me: impute.me

psychRisk: http://leobrueg.shinyapps.io/psychRisk

This presentation:  slides.com/leoo/deck-8/

 

Me: leo-brueggeman@uiowa.edu

       @LeoBman