Looking for some help in helping understand polygenic score?

Alzheimer's, cardiovascular, and other chronic diseases; biomarkers, lifestyle, supplements, drugs, and health care.
Post Reply
giftsplash
Contributor
Contributor
Posts: 127
Joined: Fri Dec 20, 2013 4:39 pm

Looking for some help in helping understand polygenic score?

Post by giftsplash »

Trying to get a handle on how to determine the risk of a disease based on ones polygenic score.

If I am in the 99% percentile for a particular disease with a 1.6 polygenic score and the odds of the disease is 2% in general population. Can I determine my odds of getting the disease?
NF52
Support Team
Support Team
Posts: 2772
Joined: Tue Oct 25, 2016 9:41 am
Location: Eastern U.S.

Re: Looking for some help in helping understand polygenic score?

Post by NF52 »

giftsplash wrote: Mon Jan 23, 2023 9:17 am Trying to get a handle on how to determine the risk of a disease based on ones polygenic score.

If I am in the 99% percentile for a particular disease with a 1.6 polygenic score and the odds of the disease is 2% in general population. Can I determine my odds of getting the disease?
Great question!

I'm not a geneticist, but if baseline risk is 2%, that means 98 out of 100 people will likely not EVER get this disease. Increasing that risk by 60%, which is what a risk of 1.6 means (a risk ratio of 1 is the same risk as anyone else; the decimal tells you what increased percentage over one.) So you MAY have a lifetime risk of 3.2 %, which is still better than 96 % chance risk of NOT having this disease. [2% x 1.6= 3.2%].

I heard a genetics counselor explain that telling someone their risk of a cancer is "60% higher than average" is often heard as "I have a 60% likelihood of this cancer". His recommendation: Always translate to "absolute risk" not "relative to the normal risk".

Here's a source that gives some real-world examples. When it mentions those in the top 5% of PRS, that would be people with a polygenic risk score of 95-99 percentile, like you for one disease.
Hence, for all but the most common cancers, PRS reveals absolute increase in cancer risk against baseline that is miniscule even at the extreme upper tail of PRS. For example, for those in the respective top 5% of PRS, lifetime risk for breast cancer is elevated from 11.8 to 19.0% (1.6-fold), lifetime risk of prostate cancer is elevated from 12.7 to 22.2% (1.75-fold) and lifetime risk of colorectal cancer is elevated from 4.6 to 6.9% (1.5-fold).
Will polygenic risk scores for cancer ever be clinically useful?

We know that risk scores can be balanced by resilience factors, especially in diseases that are known to have multiple co-factors and often great screening tools.

Hope this helps!
4/4 and still an optimist!
petri.d
Contributor
Contributor
Posts: 6
Joined: Wed Oct 13, 2021 7:23 pm

Re: Looking for some help in helping understand polygenic score?

Post by petri.d »

Hi giftsplash -

Polygenic scores are difficult to interpret for several reasons. First, for most complex diseases, only a fraction of the genetic heritability is known. For example, let's say for your disease your polygenic score is calculated from 10 variants (SNPs). Those 10 variants may only be a small fraction of genetic variants that determine your genetic risk. The remainder of genetic variants have yet to be discovered.

Second, do you know the distribution of polygenic scores for the cohort you are being compared against? For example, if the polygenic scores range from 1.4 - 1.6 then a slight decrease in your score, let's say it was 1.5 instead of 1.6, might put you in the 50% percentile (and not 99th). Because so few people have had their genome sequenced it's hard to build large cohorts to accurately compare polygenic scores.

Finally, it's important to understand the geographic composition of the group you are being compared to. Because geography plays a significant role in genetics (i.e. a person of European descent might have a very different set of variants for this particular disease) you need to know exactly who you are being compared to. To accurately interpret your polygenic score you need to be compared to a control group with matched genetic background.

I don't know about your specific disease, but in general, for the vast majority of diseases, particularly less common ones, a lot more research is needed before polygenic scores can be interpreted in a meaningful way.
giftsplash wrote: Mon Jan 23, 2023 9:17 am Trying to get a handle on how to determine the risk of a disease based on ones polygenic score.

If I am in the 99% percentile for a particular disease with a 1.6 polygenic score and the odds of the disease is 2% in general population. Can I determine my odds of getting the disease?
Post Reply