Center or fix a sample at 0 on PC1 in PCA plot

Question

[bioinf.questions]

We have recently noticed a group that is consistently able to publish PCA plots with their sample of interest seemingly aligned at 0 on PC1 or both PC1 and PC2 . In their methods they only mention that they use the R prcomp() function. We are interested to understand how they might be doing this. Attached is an example plot showing sample X at 0 on both the PC1 and PC2 axis.

Answer 1

[jared.andrews07]

My guess is that they're just picking their reference sample, determining the offsets for PC1/2 to get it to 0, and then applying that same offset to all of the other samples prior to plotting as well.

[fracarb8]

Another approach is to calculate the distance of each point to your reference (PC1), set PC1 to 0 and use the distance to position the other points.