Question: polygenic risk score
0
gravatar for maki.dur
3.4 years ago by
maki.dur0
Croatia
maki.dur0 wrote:

Hi, Recently, we got a project connected with genome genome wide association studies. We are trying to associate around 30 SNPs with aneurysm. We have in plan to calculate polygenic risk score using PRSice. The problem is that we cannot find any publicly available GWAS database as a reference.

Is there any recommendation for some possible GWAS databases or way to generate one?

I saw in PLINK resources there are some datasets(HapMap...). Does anybody have experience with that?

Thanks in advance, Marija

prs gwas • 2.7k views
ADD COMMENTlink modified 6 months ago by 29701382040 • written 3.4 years ago by maki.dur0

So you are looking for a GWAS dataset of geographically matched individuals to your population without aneurysm? It would be better to chose a title reflecting that.

ADD REPLYlink written 3.4 years ago by WouterDeCoster42k

Yes, something like that. We have sample population, but we are missing GWAS databases for reference as it's required here.

ADD REPLYlink written 3.4 years ago by maki.dur0

Obviously you need a control population. Depending on the effect sizes (relative risk) you expect, you might need a few hundred up to tens of thousands of control individuals. How many cases/patients do you have? Which population are you studying?

ADD REPLYlink written 3.4 years ago by WouterDeCoster42k

Hi, do you konw what R package that can calculate Genetic risk score ?

ADD REPLYlink written 6 months ago by 29701382040

Calculate it yourself by fitting a regression model and using the beta coefficient ('estimate'). All risk scoring models that I have seen start off with the beta coefficient, including my own.

If you want a pre-packaged program, try this https://choishingwan.github.io/PRSice/

ADD REPLYlink written 6 months ago by Kevin Blighe53k
1
gravatar for Sam
6 months ago by
Sam2.5k
New York
Sam2.5k wrote:

It might be best if you open a separate question for this. But in short, if you are looking for a pure R package for PRS, then you can use lassosum. Alternatively, you can always use PRSice and LDpred, which are the other PRS software, though they are not written in R (PRSice is technically a C++ program with an R wrapper for plotting)

ADD COMMENTlink written 6 months ago by Sam2.5k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 878 users visited in the last hour