PCA issues

[daybreak0807]

I saw in your document that you mentioned using PLINK for PCA dimensionality reduction, but I see that almost all genomes have feature vector values within ± 1 after using PLINK dimensionality reduction. Why is your data value range within ± 100? Is there any other dimensionality reduction method used?

[z4668640]

Hello, I guess the difference in values after PCA should be related to the data itself. In addition, I checked the data after PCA and found no ±1. In addition, the model supports any two-dimensional table with numeric data, in other words you can try any dimensionality reduction you are comfortable with, but PCA is more common for most people.