Problem in annotation function for visualization of points in PCA plot

Question

[Kshitij]

I am trying to replicate the Principal Component Analysis (PCA) method for a set of proteins using Bio3D in R to examine the relationship between different structures based on their equivalent residues. This is the link to the documentation of PCA in the Bio3D package - http://thegrantlab.org/bio3d_v2/tutorials/principal-component-analysis#:~:text=different%20analysis%20methods%3F-,Principal%20Component%20Analysis%20(PCA),-Following%20core%20identification When I run the final plotting step

plot(pc.xray, col=annotation[, "color"])

I face the following issue:

Error in annotation[, "color"] : 

object of type 'closure' is not subsettable

I can see the annotation function is of type of closure and cannot be used in R-4.2.0 (The version of R I'm using). I wanted to know, if there is some alternative to this function. Basically, I want to visualise different groups of points in different colours in the PCA plot as they have done in this figure (Red and green).

[LChart]

It looks like in the example they don't actually create the annotation object. They do have:

anno <- pdb.annotate(hits$pdb.id)
head(anno[, c("resolution", "ligandId", "citation")])

which produces the correct data structure (matrix). You could use this with something like

color.vec <- ifelse(resolution < 2, 'red', 'green')
plot(pc.xray, col=color.vec)

but i can't tell from that page what they are coloring.

Answer 1

[Matthias Zepper]

I think, this is an error in the documentation / tutorial.

annotation[, "color"]

is the notation one would use to select a column named color from a data.frame called annotation.

Because this data.frame is never created throughout the tutorial, annotation then resolves to a function of the same name. You can try this with any of the default functions, for example which():

> which[,"color"]
Error in which[, "color"] : object of type 'closure' is not subsettable

> which
function (x, arr.ind = FALSE, useNames = TRUE) 
{
    wh <- .Internal(which(x))
    if (isTRUE(arr.ind) && !is.null(d <- dim(x))) 
        arrayInd(wh, d, dimnames(x), useNames = useNames)
    else wh
}
<bytecode: 0x141a7f860>
<environment: namespace:base>

But of course we can treat which like an ordinary variable name and assign a data.frame to it. This is done here for demonstration purposes only and generally a bad practise.

> which <- data.frame(color=1)
> which[,"color"]
[1] 1
> which
  color
1     1

Now, it is possible to use which[,"color"].

To fix your plot, you have to provide a vector of color values to col=.