Hi all,
I am using plink to adjust for population structure and have two questions, one straightforward implementation one conceptual.
When using plink 2.0 and --pca, can I (should I) directly use the columns of my resulting my_res.eigenvec in a model as adjustment variables (if relevant, my modeling will be outside of plink), or do I need to transform the output in some way? I think I directly use the columns (or some subset of the columns) as adjustment variables, but I am not confident. I think my confusion originates from the fact that these are labeled as eigenvectors (and also a weak understanding of PCA, despite lots of reading).
In detail, are the columns of .eigenvec actually eigenvectors (I don't think so) or are they principal components (or principal components scores? not sure if there is a difference)? How from my output or the properties of eigenvectors/PCs should I be able to tell whether these are eigenvectors or PCs myself?
Thank you!
