Massive AD GWAS! YEAH!!!

Insights and discussion from the cutting edge with reference to journal articles and other research papers.
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Massive AD GWAS! YEAH!!!

Post by J11 »

Above is a first try at formatting Table 1a and b. The pdf conversion did not work out properly and they interwove the two tables and some of the characters were not OCRed correctly. It's a start. This table is our best insight yet into the genetics of AD.
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Massive AD GWAS! YEAH!!!

Post by J11 »

SusanJ, alzforum recently reported on a clinical grade polygenic AD test.

"As ever-larger genome-wide association studies make clear, a person’s genetic risk for Alzheimer’s disease is driven by the cumulative effects of myriad variations in their genes. The public will now have access to a test that sums up that risk from a few drops of spit. The genoSCORETM-LAB, a polygenic test developed by Cytox Ltd. in Manchester, England, U.K., uses a custom chip to tally genotypes at thousands of genomic positions."

At some point one suspects that this will convert down to a simply genochip upload and online score.
Given that the results from Table 1a and b are published this could be done DIY. Of course, I suspect
many would just be as happy to pay a modest fee and have it done for them.

https://www.alzforum.org/news/research- ... imers-risk


There is of course a nearly unlimited amount of interesting other polygenic scores that are available in the literature.
This is where the R code could be handy so that one could calculate scores for oneself.
All of this genetic information is inevitably leading us to a world highly highly aware of these polygenic scores and
the likely consequences of a genetically selected society.
Last edited by J11 on Sat Mar 13, 2021 1:13 pm, edited 1 time in total.
User avatar
SusanJ
Senior Contributor
Senior Contributor
Posts: 3058
Joined: Wed Oct 30, 2013 7:33 am
Location: Western Colorado

Re: Massive AD GWAS! YEAH!!!

Post by SusanJ »

J11 wrote:SusanJ, alzforum recently reported on a clinical grade polygenic AD test
.

Thanks, I'll take a look. I'm definitely in the modest fee camp.
Jgreg80
Contributor
Contributor
Posts: 12
Joined: Wed Feb 13, 2019 12:37 pm

Re: Massive AD GWAS! YEAH!!!

Post by Jgreg80 »

Appreciate it, J11

Unfortunately, neither 23andMe v4 nor Ancestry v2 genotyped the two new ones with large ORs (SORT1, NCK2).
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Massive AD GWAS! YEAH!!!

Post by J11 »

SusanJ, the paradox of the modern economy is that what we value the most has the least (marginal) cost. The post-economic world has arrived. Alzheimer's GWAS have cost millions and millions of dollars and yet once this information is in hand there is no great cost involved in finding the polygenic score. There is no obvious way that the funders can be given a direct return on their investment. Basic science often does not translate into a marketable. Probably the best way of achieving the return is to simply allow it to be open data. Freeing the data would allow the overwhelming benefit to society and others of early detection and treatment.

The recent GWAS article found over 40 new AD loci and is starting to provide a very good view of the genetics of Alzheimer's. Yet, these SNPs were all posted in the article and with a full genome in hand ( current price ~$300) one could search through and locate all of these variants in probably an hour or two. Marginal cost (excluding time value approaches zero). However, for those on forum this information is of near crucial importance. It is quite possible that this polygenic score could even provide reassurance to an e44 that they were not actually in the expected e44 high risk group.
User avatar
karelena
Contributor
Contributor
Posts: 81
Joined: Thu Feb 05, 2015 3:11 pm

Re: Massive AD GWAS! YEAH!!!

Post by karelena »

How is it possible to know the effect of someone's unique combination of AD related genes on their risk of AD? How do they come up with the polygenic test "score?" I am wondering if the test is meaningful for us now, or if whatever algorithms they are using need to be validated over many years with actual clinical data (who gets AD and who doesn't).
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Massive AD GWAS! YEAH!!!

Post by J11 »

karelena, what the GWAS research has found is that all that is needed to find a polygenic score is to add up the number of risk alleles weighted by their betas. Strangely, such a seemingly invalid approach actually captures the risk of most human traits. The idea here is that biology is a super-complex system and so that for non-dominant illnesses (e.g. Alzheimer's, i.e., PSEN etc.) much of the risk is polygenic or even omnigenic (which is to say essentially that someone's entire genome might contribute in very tiny ways to disease risk).

The current AD polygenic score has now achieved a truly polygenic (~100 loci) character and it is reaching the point of being clinically valid. By themselves each of the recently described SNPs have very minimal influence, though when summed up they can have substantial bearing on risk. The polygenic AD product described recently on alzforum is quite impressive.

People will have clear warning before they are coping with symptomatic AD. This is remarkable because it has not really been that long that apoe4 testing has even been available. In that previous time there was considerable doubt whether it should even be available at all. This forum would not exist without such testing and the dementia consciousness that has developed would also not be present. Basically, the e44s on forum would have realized that they were at dementia risk when they began declining into the symptomatic stage. I doubt many on forum would actually choose such a path now. It is fairly clear that by knowing genotype one can make AD a life focus and create the social- political - economic conditions through our collective life force to overcome our dementia challenge.

Others without epsilon 4 can also be at high risk and this would also have gone unnoticed without the recent polygenic test. Given the highly polygenic nature of AD what the lack of a polygenic score meant was that a substantial number of typical ordinary people would wake up one day and out of the blue be faced with the mystery of AD without any family history of dementia. This happens because with a truly polygenic illness anyone on a bad day could wind up with a bad roll of the dice. If this bad roll happened at the time of genetic recombination, then AD could by a mere strike of bad luck afflict anyone.

Hopefully, though even more and bigger GWAS will be done and the entire AD genome will unlock. I am surprised this has not been more of a priority. Within my family there appears to be a rare dominant AD risk variant that has not been well reported in the genetics literature. Yet, the cost to society of letting these family level genotypes to be undiscovered is large. Simply sampling from memory clinics and then tracing ancestors on GEDmatch could rapidly fill in many of the gaps in knowledge, though for whatever reason this has largely not been attempted at scale to date.
Last edited by J11 on Sun Mar 21, 2021 12:45 pm, edited 1 time in total.
NF52
Support Team
Support Team
Posts: 2772
Joined: Tue Oct 25, 2016 9:41 am
Location: Eastern U.S.

Re: Massive AD GWAS! YEAH!!!

Post by NF52 »

karelena wrote:How is it possible to know the effect of someone's unique combination of AD related genes on their risk of AD? How do they come up with the polygenic test "score?" I am wondering if the test is meaningful for us now, or if whatever algorithms they are using need to be validated over many years with actual clinical data (who gets AD and who doesn't).
Hi karelena,

Great question in a still emergingfield! Here's an excerpt from a March 4 article in Alz Forum, which focuses on recent AD research and tends to be a little "wonky" in vocabulary, so I've added some "translations" in brackets: [url=https://www.alzforum.org/news/research- ... -39771Just What the Doctor Ordered? Polygenic Test Gauges Your Alzheimer’s Risk[/url]
Several iterations of polygenic risk score have been developed over the years. While some gauge genetic risk by weighing the contributions of variants that rose to genome-wide significance [statistically significant in isolation]in GWAS [genome-wide association studies], others take the broader approach of incorporating thousands of variants scattered across the genome even if they didn’t reach that statistical threshold.
Using data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), [the company]Cytox recently reported that its polygenic risk score predicted subsequent cognitive decline in people who were cognitively normal or who had mild cognitive impairment (Daunt et al., 2021). The genetic test rivaled, but did not outperform, the CSF p-tau/Aβ42 [spinal tap measurements of tau and amyloid beta] ratio in predicting cognitive decline over four years in people with MCI [mild cognitive impairment]. It trended toward predicting decline in people who were cognitively normal, but there were too few participants in this category to reach statistical significance. The test, which includes APOE genotype, was predictive regardless of whether a person carried an ApoE4 allele...the genetic risk score can be gleaned from a saliva sample donated by the participant at home... [the] GenoScore is not intended to replace fluid biomarkers, but to be used in conjunction with them, or as the initial part of a screening pipeline.

Genoscore has received the ...approval that is required for marketing tests in Europe, and the company expects to receive a CLIA designation as a laboratory developed test (LDT) in the United States before May, Pither told Alzforum. These approvals recognize that products meet manufacturing, technical, and health standards, i.e., that Genoscore accurately and reliably genotypes samples. Even so, Pither said, further testing will need to validate that the polygenic score predicts AD risk in the general population, and to show that knowing one’s genetic AD risk has health benefits.
That last sentence seems crucial--these tests have been used on small samples of people with normal cognition, and it's unclear what percentage of those had ApoE4. All would typically need to be followed for4 years or more like current clinical trials to validate how accurate they are in predicting a future diagnosis.
4/4 and still an optimist!
User avatar
karelena
Contributor
Contributor
Posts: 81
Joined: Thu Feb 05, 2015 3:11 pm

Re: Massive AD GWAS! YEAH!!!

Post by karelena »

NF52 wrote: That last sentence seems crucial--these tests have been used on small samples of people with normal cognition, and it's unclear what percentage of those had ApoE4. All would typically need to be followed for4 years or more like current clinical trials to validate how accurate they are in predicting a future diagnosis.
It's going to take a lot longer than 4 years to see if normal cognition subjects end up with AD. Decades.
Post Reply