How to describe bi-plot PCA figure?

Question

[Emy]

Hello everyone,

I have this bi-plot PCA in my manuscript. I submitted this manuscript in a journal and now one of the reviewers has said “the authors do not identify clearly what are the PC1 and PC2 components”. I really couldn’t figure out if the reviewer means the PC1 and PC2 which are on axes x and y?! If so, I have never seen any literature to explain the axes and always the programs give us PCA figures with x and y axes showing PC1 on axis x and PC2 on axis y and a number in %. Or I’m wondering maybe the reviewer means to identify graphic 1 and graphic 2 of this figure. Could you please let me know your thoughts about this question regarding this figure?

‏please feel free to reach me with any more questions and information. With regards

Answer 1

[Mensur Dlakic]

I think what the reviewer means is that there is some property in your data, either simple or composite, that is equivalent to PC1, and similarly to PC2. It may or may not be obvious what those properties are, but the reviewer is asking if you considered any possibilities to explain them.

In the simplest case, two of your original data features are (mostly) aligned with your PCs and can explain most of the variance. In the more complex case - and likely more realistic - there is some physical or mathematical property that correlates with your data features that can explain most of the variance, and it would be up to you to find it.