PLINK - Obtaining polygenic score

Question

[amalia]

Hi,

I am trying to obtain in PLINK the polygenic scores of a sample of individuals. I simulated a very simple example to see whether I could make it work, but it seems that it is not.

I have simulated data for 3000 individuals and 2 SNPs. I created a file .raw that has three columns: SNPid, Reference allele, and score. In the score column I put the log of odds ratio for significantly associated SNPs. In the reference allele column I just randomly set everything to "A", but I do not know whether this is correct, and possibly this is why I am getting an error afterwards. I ran the --score command in plink and I obtain a file where the "score" column is all "0".

Can anyone help?

What do I have to put for "reference allele" if my data is simulated?

Thanks

Answer 1

[NB]

So, Your score file should have lines looking like

SNP1 allele_effected Score

The second column is not a 'reference allele', it is more of the affected allele for that SNP that contributes to the score.

There are many tutorials out there, for example this one

[amalia]

Thank you!

I have a more basic question then. I simulated data for 3000 individuals and 2 SNPs. For the 2 SNPs I simulated two columns for each, one indicating the allele on the chromosome inherited from the mother and another one from the father. I have filled up these two columns with a lot of 1 and 2's. However, In my .raw file I need to state which is the allele_effected in terms of "G-C-T-A". How do I know this from my simulation?

What do these 1 and 2's that I have simulated mean?

Can I put in my .raw file a 1 or a 2 as allele_effected?