Polygenic info used to identify risk of AD

Insights and discussion from the cutting edge with reference to journal articles and other research papers.
xingxu
Contributor
Contributor
Posts: 17
Joined: Thu Oct 04, 2018 9:56 am

Re: Polygenic info used to identify risk of AD

Post by xingxu »

I will reply to other questions shortly.
Disclaimer: I am one of the co-founders of Dash Genomics.
xingxu
Contributor
Contributor
Posts: 17
Joined: Thu Oct 04, 2018 9:56 am

Re: Polygenic info used to identify risk of AD

Post by xingxu »

MarcR’s understanding is correct. PHS 100 only means that you are probably in the highest risk strata. A PHS percentile presents a GROUP of people who share the similar risk for AD.

@circular
These early-onset Alzheimer's disease genes (e.g. PSEN) are excluded from the analysis. The model is only for Late-Onset Alzheimer's Disease.
MarcR wrote:
circular wrote:I really wonder how they can have a 100 percentile unless they have the deterministic PSEN and APP genes in there?
I think the PHS is a relative metric, not an indicator of absolute risk. "100" means maximum risk relative to other humans, not 100% risk.

AZ50 is the absolute risk metric. An AZ50 of 80 indicates that we should expect 50% of people with similar genetics to have LOAD by age 80.
Disclaimer: I am one of the co-founders of Dash Genomics.
xingxu
Contributor
Contributor
Posts: 17
Joined: Thu Oct 04, 2018 9:56 am

Re: Polygenic info used to identify risk of AD

Post by xingxu »

@MarcR

Great question. We thought about how to present the comparison between the individual's risk and the average risk of the population with APOE-ε4 genotypes. We actually already have the average cumulative risk for APOE-ε4+ population (people who have 1 copy or 2 copies of APOE-e4) in the backend. We are in discussion to separate the average risk of APOE-ε4+ into two more precise groups of APOE-ε4 (heterozygote, 1 copy) and APOE-ε4 (homozygote, 2 copies). We will make this information available when it is ready.

MarcR wrote:
xingxu wrote:It’s worth noting that in the context of the polygenic score, those high-impact loci does NOT necessarily mean causation. They only suggest a high correlation between genotypes of loci and phenotypes (the risk of Alzheimer’s disease in this case).
My primary interest in seeing the results of other forum members is to get a feel for the degree of variation in PHS and AZ50 among other heterozygous APOE-ε4 carriers. The original paper implied that risk would vary widely - some heterozygous carriers might have risk below that of the average ε3/ε3, and others might have risk approaching that of typical homozygotes. With only two data points, I don't have much perspective on my genetic luck. Do I have more or less risk than the average heterozygote? How much more or less? How does my luck compare with that of the average homozygote?

From an actionability standpoint, just how motivated should I be to take extraordinary measures to maintain and protect my cognition?

Xing, I wonder if Dash has any data to share with us on this topic?
Disclaimer: I am one of the co-founders of Dash Genomics.
User avatar
MarcR
Mod
Mod
Posts: 2017
Joined: Wed Mar 05, 2014 8:28 pm
Location: Sammamish, Washington, US

Re: Polygenic info used to identify risk of AD

Post by MarcR »

xingxu wrote:We are in discussion to separate the average risk of APOE-ε4+ into two more precise groups of APOE-ε4 (heterozygote, 1 copy) and APOE-ε4 (homozygote, 2 copies). We will make this information available when it is ready.
Thanks, Xing. I'm glad to hear this and look forward to reviewing the information when it's ready.
xingxu
Contributor
Contributor
Posts: 17
Joined: Thu Oct 04, 2018 9:56 am

Re: Polygenic info used to identify risk of AD

Post by xingxu »

The analysis takes 600,000 loci in the genome into the consideration for Alzheimer's risk. It is by far the most comprehensive analysis on the market. It includes loci that are thought to be protective and loci that are associated with risk. In certain cases, a person who has a copy of APOE-e4 may have the similar risk as an e3/e3 person because of protective loci.

We were excited to make available this research for a few reasons:
1. It provides a more comprehensive view of the genetic risk for Alzheimer's disease, which goes beyond APOE. See my answer to your question about the protective loci.
2. The analysis is quantitative. For example, we designed the concept of AZ50 age which provides a quantitative measure about the risk. People with the same APOE subtype might have different risk profiles because of other loci in their genomes and this analysis provides a basis for comparison.
3. We believe "Power in Knowledge". As we all know, the brain begins changing 10-20 years before symptoms of Alzheimer's disease onset. The age of onset for an individual with low risk versus one with high risk can vary by as much as 15 years or more based on research conducted by UCSF and UCSD. When the symptoms onset, it might be too late to respond to the risk. We already see an ongoing paradigm shift of Alzheimer's disease treatment from treating the disease to preventing the disease.
By providing the knowledge for people to understand the risk in a quantitative, comprehensive and intuitive way, we hope this knowledge can motivate people to make more informed lifestyle changes that might help fight the odds.
Julie G wrote:
Have you done it? I'm very interested in a virtual meet up.
LOL, I'm not sure why a 4/4 (with other putative genes) would pay $99 to learn I have a very high risk.

Xing, perhaps you can weigh in here. Does your retail algorithm take into account any protective genes? For instance, I recently ran one of Rhonda Patrick's gene assessment tools and was pleasantly surprised to learn that I carry ten longevity snips. Would that temper my score?

As long as we have a few others (minimum 6?) who are interested, I'd be happy to set up a meeting :D.
Disclaimer: I am one of the co-founders of Dash Genomics.
User avatar
MarcR
Mod
Mod
Posts: 2017
Joined: Wed Mar 05, 2014 8:28 pm
Location: Sammamish, Washington, US

Re: Polygenic info used to identify risk of AD

Post by MarcR »

Xing, those of us who follow this topic have three data points:
  1. 3/4: PHS=88, AZ50=86
  2. 3/4: PHS=89, AZ50=86
  3. 4/4: PHS=100, AZ50=80
My working assumption is that 599,998 of the loci you're examining have little effect and/or are strongly correlated with APOE genotype and that the vast majority of 3/4's will be in the upper 80's, and the vast majority of 4/4's will be 100 or in the upper 90's. Until we understand the distribution of scores by APOE genotype - 2/2, 2/3, 2/4, 3/3, 3/4, 4/4 - why should we assume otherwise?
circular
Senior Contributor
Senior Contributor
Posts: 5565
Joined: Sun Nov 03, 2013 10:43 am

Re: Polygenic info used to identify risk of AD

Post by circular »

Thanks Xing for sharing your additional thoughts. One thing that would really help me would be a video explanation with graphics to explain the science behind how the PHS was developed and how it compares to other work. I just don't have a good grasp of the process and visuals might really help.

In the waaaaaay distant future maybe we'll be able to input our lifestyle factors for an epigenetically modified score :D
ApoE 3/4 > Thanks in advance for any responses made to my posts.
User avatar
Julie G
Mod
Mod
Posts: 9187
Joined: Sat Oct 26, 2013 6:36 pm

Re: Polygenic info used to identify risk of AD

Post by Julie G »

In the waaaaaay distant future maybe we'll be able to input our lifestyle factors for an epigenetically modified score :D
Yes, please!
MyDestinyisfromGod
Contributor
Contributor
Posts: 13
Joined: Thu Nov 01, 2018 5:05 am

Re: Polygenic info used to identify risk of AD

Post by MyDestinyisfromGod »

Hi my name is Lynn. I am 53. I just received my results from DASH. I am thankful for the people that have invested time and money to help shed light on the research and things we can do. I have an AZ50 of 86. The risk goes up almost 10 percent each year until by the age of 90, it just says <95. The one thing that was a bit confusing to me was the instantaneous risk. If you look at age 95 it shows 26%. That seems a bit irrelevant if it is next to certain that I’ll have it by age 90 anyway. So I thought I might be missing something important with that.

Secondly, I wondered if the results would be different if my Dad who had Alz (died 3 years ago) had early onset. If I inherited something from him, and he did have early onset, would that mean my AZ50 would be younger (and the rest correspondingly so)? The best recollection my brother and sister have, I think he showed really clear symptoms at age 63. He looked at a frisbee at a birthday party at 62 and didn’t know what it was. He had played frisbee with us as kids. When we explained it, he tried to throw it and it just went straight to the ground. So it was really full on at that point. Because we are all grown and gone, we don’t know when the symptoms began. My Mom gets her feelings hurt if we ask. She feels it’s somehow shameful, and wants to think (despite the diagnosis) that it was probably the medicine he was on making him forgetful. I called to find out how my Dad’s brother’s died and she talked about it, but then told my brother that I was apparently intent on hurting her.

When I questioned a study at IU about if he could have had early onset, the lady I talked to said that there would have been several family members with it and it really should be before the age of 60 to qualify as early onset. He had one brother with severe dementia, but not Alz, and another one with a brain tumor, as well as a nephew (from another brother) that died of brain cancer. So there are brain related things, but not actually Alzheimer’s. Lately everything I see says it’s early if the onset is before 65 though.

Sorry if I’m straying off the subject. I’ll check on some other forum to find out more about if he could have had early onset. I just want to know if those facts could skew my results. Also, I hope I didn’t post this twice. I am new to the forum and typed this all out last night. But with distractions, it took me long enough that when I hit submit, it made me log in again and the post was gone. I know a moderator needs to approve it first, but when I didn’t see anything today. I thought maybe it had actually been deleted.
MyDestinyisFromGod
User avatar
MarcR
Mod
Mod
Posts: 2017
Joined: Wed Mar 05, 2014 8:28 pm
Location: Sammamish, Washington, US

Re: Polygenic info used to identify risk of AD

Post by MarcR »

MyDestinyisfromGod wrote:I have an AZ50 of 86.
Lynn, would you be willing to share your APOE genotype (from 23andMe) and your Polygenic Hazard Score percentile (from Dash)?
The risk goes up almost 10 percent each year until by the age of 90, it just says <95. The one thing that was a bit confusing to me was the instantaneous risk. If you look at age 95 it shows 26%. That seems a bit irrelevant if it is next to certain that I’ll have it by age 90 anyway. So I thought I might be missing something important with that.
I interpret that as the risk of your developing AD in the coming year if you are 94 and do not already have it.
Secondly, I wondered if the results would be different if my Dad who had Alz (died 3 years ago) had early onset.
I think the model Dash implemented is for Late Onset Alzheimer's Disease (LOAD) only. I think the early onset form is classified as a different disease, and it seems to be tied to rare polymorphisms in the PSEN1 and PSEN2 genes. For $12 you can feed your 23andMe data to Promethease and get a cool report comparing your genes to all the research in SNPedia - I'm sure you'll see immediately if you have any issues there. Alternatively, you can visit SNPedia, search for PSEN1 and PSEN2, and manually search your 23andMe file for any problematic SNPs.
Post Reply