With only 1200 affecteds they found 6 new genes and confirmed another?
They needed nearly 100,000 in IGAP to find 22!
https://www.biorxiv.org/content/early/2018/01/09/242545
New GWAS design identifies 6 new AD genes!
Re: New GWAS design identifies 6 new AD genes!
Interesting. Looks like a pre-print of an interesting paper. I'm interested to learn more about the relative contributions or the genes they found, compared to those we know about. I'll have to get into the math of this paper and compare to some others. It would make it super-easy if "e4" risks were a standard measure and everything else was compared to that.....like this COMT gene conveys an increased risk of 0.3 "e4 units", or something like that. I'm sure that math would be quite a challenge but it sure would be helpful.
Re: New GWAS design identifies 6 new AD genes!
Tom, this is GREAT!
They spent a huge amount of money with IGAP and people can get tired with all these expenses and not a great deal of results.
With this new research approach they used 100 times fewer people and obtained a third as many genes.
Not bad!
I would say that this represents a compelling research strategy from the point of view of maximum value for the dollar.
How could additional such studies not be funded?
This also highlights my suggestion that perhaps we could get online and do this ourselves.
The family pedigrees that they show near the end of the article are not dynasties by any stretch of the imagination.
They are only showing a few affecteds in most families.
The researchers did not appear to have taken advantage of GEDmatch to drive home their results.
It is a very large mystery to me why this was not done.
They finally report that they are confident that some of the reported genes are true hits, though they are not confident that they all are. If they got onto GEDmatch, they could have nailed it. Instead of having possibly only a small number with any given rare AD genotype they probably would have soon been able to find dozens in the extended with it. Once you hit a branch of the family tree, affecteds will soon be coming out of the woodwork everywhere!
If they had narrowed down their search space to a single gene, then why not find distant cousins of their families who shared these variants?
I have found interesting variants in the genes they found and I might just go to GEDmatch to find family
on the4se stretches. For example, I found a missense variant in MAS1L in our exome file with a massive quality score of 1500 and a MAF of 0.001. Mutation called it a polymorphism, though you never know. How worried would you be?
Also found a missense variant in CPAMD8 good quality which was somewhat less rare at 0.01. How is it possible that I could plausibly outcompete a research team with all their inherent advantages?
It is very disappointing that they did not include the SNPs that they found in the study! Why not share the information? Isn't that what science is supposed to be all about? If they reported a SNP that we had in our exome file with a MAF of 0.001, I would certainly take note. It would be even all the more helpful if genetic researchers could link up with snpedia and use the power of Promethease to bring the research to the people! Why does this all have to be so cumbersome?
They spent a huge amount of money with IGAP and people can get tired with all these expenses and not a great deal of results.
With this new research approach they used 100 times fewer people and obtained a third as many genes.
Not bad!
I would say that this represents a compelling research strategy from the point of view of maximum value for the dollar.
How could additional such studies not be funded?
This also highlights my suggestion that perhaps we could get online and do this ourselves.
The family pedigrees that they show near the end of the article are not dynasties by any stretch of the imagination.
They are only showing a few affecteds in most families.
The researchers did not appear to have taken advantage of GEDmatch to drive home their results.
It is a very large mystery to me why this was not done.
They finally report that they are confident that some of the reported genes are true hits, though they are not confident that they all are. If they got onto GEDmatch, they could have nailed it. Instead of having possibly only a small number with any given rare AD genotype they probably would have soon been able to find dozens in the extended with it. Once you hit a branch of the family tree, affecteds will soon be coming out of the woodwork everywhere!
If they had narrowed down their search space to a single gene, then why not find distant cousins of their families who shared these variants?
I have found interesting variants in the genes they found and I might just go to GEDmatch to find family
on the4se stretches. For example, I found a missense variant in MAS1L in our exome file with a massive quality score of 1500 and a MAF of 0.001. Mutation called it a polymorphism, though you never know. How worried would you be?
Also found a missense variant in CPAMD8 good quality which was somewhat less rare at 0.01. How is it possible that I could plausibly outcompete a research team with all their inherent advantages?
It is very disappointing that they did not include the SNPs that they found in the study! Why not share the information? Isn't that what science is supposed to be all about? If they reported a SNP that we had in our exome file with a MAF of 0.001, I would certainly take note. It would be even all the more helpful if genetic researchers could link up with snpedia and use the power of Promethease to bring the research to the people! Why does this all have to be so cumbersome?
