In the above R command, the "t" function returns the
transpose of its argument, and the "cmdscale" function returns
a list object whose "points" element contains the PCoA
coordinates.- Merge the PCoA results with the sample metadata and visualize
PCoA results as a scatter plot. Each of these steps can be
performed in a single R command as shown below (with exam-
ple output):
merged_pcoa_data <- data.frame(sample_data_merged, pcoa_results, )
head(merged_pcoa_data, n=15)
condition species X1 X2
TCC.1 cancer dog -5.20 -1.01
TCC.2 cancer dog -3.97 -1.38
TCC.3 cancer dog -2.31 5.53
TCC.4 cancer dog 3.65 3.43
TCC.5 cancer dog -2.70 2.01
TCC.6 cancer dog 0.27 4.89
TCC.7 cancer dog 3.66 1.68
normal.1 normal dog -1.11 -4.41
normal.2 normal dog 1.73 -5.33
normal.3 normal dog 3.94 -5.44
UCCB_1 cancer human 2.23 2.44
UCCB_2 cancer human -5.97 5.65
UCCB_3 cancer human -7.65 5.80
UCCB_4 cancer human -2.98 6.34
UCCB_5 cancer human -9.29 -5.45
dim(merged_pcoa_data)
[1] 443 4
library(ggplot2)
ggplot_res <- ggplot(data=merged_pcoa_data) +
geom_point(aes(x=X1, y=X2, colour=condition, shape=species, alpha=species), size=4) +
guides(alpha=FALSE, size=FALSE) + theme(text=element_text(size=30)) +
scale_alpha_manual(values = c(1.0, 0.3))
print(ggplot_res)
The PCoA co-visualization of the human and dog bladder
datasets is shown in Fig.2 (seeNote 3). In the above R
command, the function "ggplot" sets the data source for the
plot, the function "geom_point" creates a scatter plot; the
"theme" function is used to set the point size for the text labels
in the plot; and the function "scale_alpha_manual" defines the
two alpha transparency levels to be assigned to the two values
for the "species" column in the data frame.- For each species, and for each gene that is expressed above
background, compute an intraspecies log 2 expression ratio
andp-value (for the test of equal means of the log 2 expression
levels of the two sample groups) based on a negative binomial
distribution-based count model:
Cross-Species RNA-Seq Analysis 301