How to identify which pdbs are which on PCA and NMA plotting

[Former user]

Hi everyone, I was just wondering if there is an easy way to show which PDB numbers are showing at which point in my PCA plot.

I've tried using the command: "identify(pc.xray$z[,1:2], labels=basename.pdb(pdbs$id))" The command above does work, however, my PCA graph has clustered points on it, so it is hard to tell which PDB numbers show up where because it just makes a whole mess after clicking on each point to identify.

Please feel free to reach out and let me know.

[Xinqiu Yao]

Hi,

‌

You might want to try the ggplot2 and ggrepel packages for this purpose. The following example can label all structures automatically in a “self-repulsive” manner and so labels are not overlapped:

# An example PCA
library(bio3d)
attach(transducin)
pc <- pca(pdbs, fit=TRUE)

# First, create a data frame for plotting.
dat <- data.frame(x=pc$z[, 1], y=pc$z[, 2], 
   lab=pdbs$id, col=annotation[, "color"])

# Load packages. Need to install first if absent.
library(ggplot2)
library(ggrepel)

# Plot
ggplot(dat, aes(x=x, y=y, color=col, label=lab)) +
   geom_point() +
   geom_text_repel(size=3)

‌

It may not always perfect, but after a few tweaking and twisting, you may get a pretty descent result. See manuals of both packages for more detail.

‌

Hope it helps.

[Daniel Lim]

I’ll check them out, thank you for the recommendation.