Polygenic Risk Scores

Question 1

How are polygenic risk scores calculated?

Answer 1

Sum of risk alleles weighted by GWAS effect size

Question 2

What steps need to be performed in model fitting?

Answer 2

Linkage disequilibrium, identify significance levels for true association, adjust for population architecture (ancestry)

Question 3

What are limitations inherent to polygenic risk scores?

Answer 3

Cannot speak to underlying biology, only association
Not diagnostic
Only makes sense in population context
Risk bounded by amount of time
Variable methods - can have different results

Question 4

What are current limitations with polygenic risk scores that may change with time?

Answer 4

Genotyping is largely assay based (bad for diversity and discovery)
Ancestry adjustment is optimized in White, European groups
PRS algorithms will continue to mature (versioning)

Question 5

What are potential uses of polygenic risk scores?

Answer 5

Risk prediction: wanting to understand whether we can find people at low/high risk and stratify them
Therapeutic response or intervention: who benefits the most from treatment?
Prognostic indicator: can we predict disease course?

Question 6

What are advantages of using polygenic risk scores as a screening tool?

Answer 6

One time physical test
screens for multiple conditions
Can do this before disease onset
Can control return of results to actionable times
Returns a single value
Stable throughout lifespan
May be a more accurate estimation of heritable risk than conventional risk factors or family history