All the nuts and bolts (polygenic scores&factor analysis + correlation with phenotype, MCV) of my polygenic method at work on height increasing alleles, providing evidence that populations were subject to different (sexual?)selection pressure on stature:

http://f1000research.com/articles/4-15/v1

All the nuts and bolts (polygenic scores&factor analysis + correlation with phenotype, MCV) of my polygenic method at work on height increasing alleles, providing evidence that populations were subject to different (sexual?)selection pressure on stature:

http://f1000research.com/articles/4-15/v1

Davide,

did you ever check if including nominal SNPs altered your cog pop results -- for example using those listing in the CHARGE consortium analysis suppl.

i guess no one is going to do a proper admix x g study -- though not for lack of data. a couple of the CHARGE consortium samples were sub samples of ones in which admix mapping analyses were conducted e.g.,
the Atherosclerosis Risk Study and in which admix x educ associations were found.

All the nuts and bolts (polygenic scores&factor analysis + correlation with phenotype, MCV) of my polygenic method at work on height increasing alleles, providing evidence that populations were subject to different (sexual?)selection pressure on stature:

http://f1000research.com/articles/4-15/v1

Davide,

did you ever check if including nominal SNPs altered your cog pop results -- for example using those listing in the CHARGE consortium analysis suppl.

i guess no one is going to do a proper admix x g study -- though not for lack of data. a couple of the CHARGE consortium samples were sub samples of ones in which admix mapping analyses were conducted e.g.,
the Atherosclerosis Risk Study and in which admix x educ associations were found.

The CHARGE consortium basically provided 3 genome-wide significant hits NOT in LD (13 if including those in LD).
One SNP in the CHARGE study essentially replicates one of Rietveld's hits, as they are in strong LD (r>0.9). The remaining significant hits are conflicting as one loads negatively and the other one positively on the gi factor. Emil and I are working on a paper with these results.
I checked the SNPs below the significant threshold in the CHARGE study. No signal of selection nor population structure was detected. This is very odd, it seemed entirely a result of random noise as the generated factors didn't even respect the genome-wide population relationships.

All the nuts and bolts (polygenic scores&factor analysis + correlation with phenotype, MCV) of my polygenic method at work on height increasing alleles, providing evidence that populations were subject to different (sexual?)selection pressure on stature:

http://f1000research.com/articles/4-15/v1

Davide,

did you ever check if including nominal SNPs altered your cog pop results -- for example using those listing in the CHARGE consortium analysis suppl.

i guess no one is going to do a proper admix x g study -- though not for lack of data. a couple of the CHARGE consortium samples were sub samples of ones in which admix mapping analyses were conducted e.g.,
the Atherosclerosis Risk Study and in which admix x educ associations were found.

We/you/I should write them a polite email and see if they will run the analysis for us.

Be sure not to mention that we're interested in the genetic model. When I ask for stuff like this, I always make it seem like I am interested in the colorism hypothesis, which I also am so there is no lying, only misdirection.

Unless they're dumb they'll glance that the aim of the study is. But it's worth a try.

I checked the SNPs below the significant threshold in the CHARGE study. No signal of selection nor population structure was detected. This is very odd, it seemed entirely a result of random noise as the generated factors didn't even respect the genome-wide population relationships.

Sounds ominous.

I checked the SNPs below the significant threshold in the CHARGE study. No signal of selection nor population structure was detected. This is very odd, it seemed entirely a result of random noise as the generated factors didn't even respect the genome-wide population relationships.

Sounds ominous.

Maybe but I am optimistic.I think those below genome-wide significance are not genuine hits or their phenotypic effect is too weak to undergo selective pressure.

Maybe but I am optimistic.I think those below genome-wide significance are not genuine hits or their phenotypic effect is too weak to undergo selective pressure.

Well, at very least, they show that the height results are not tautological. One need not find evidence of selection given your method.

I checked the SNPs below the significant threshold in the CHARGE study. No signal of selection nor population structure was detected. This is very odd, it seemed entirely a result of random noise as the generated factors didn't even respect the genome-wide population relationships.

Could you post the SNP pop frequency correlation matrix sometime?

Could you post the SNP pop frequency correlation matrix sometime?

Also, did you try replicated nominals i.e., excluding novel nominals?

I checked the SNPs below the significant threshold in the CHARGE study. No signal of selection nor population structure was detected. This is very odd, it seemed entirely a result of random noise as the generated factors didn't even respect the genome-wide population relationships.

Could you post the SNP pop frequency correlation matrix sometime?

I can just post the allele frequencies and you can calculate the correlation matrix. Which SNPs do you want? The top 6 from Rietveld and the CHARGE study?Freqs for top 6 are here: https://docs.google.com/spreadsheets/d/1VuRF5_YiTj50XVJeI5WeVrCGadY-EjeefR1FtWrb96o/edit?usp=sharing

Well, at very least, they show that the height results are not tautological. One need not find evidence of selection given your method.

Can you explain what you mean by "tautological" in this context?

Can you explain what you mean by "tautological" in this context?

predetermine by the method

i knew they weren't, but I have heard this argument