Click the button to launch a Binder R session and tinker with all the
bar plots or explore the code in the cloud. Navigate to the
directory where there is a subdirectory for each of the four examples,
all identified by the first or last author.
Venn diagrams are a type of figure that I see a lot in the literature. While Venn diagrams are a great way to conceptualize the goal of looking for shared or unique transcriptional responses to experimental manipulations, I don’t think they are the best way to vizualize empirical data.
I searched for “Venn” in Garrett Grolemund & Hadley Wickham’s R for
Data Science book and in Claus Wilke’s
Fundamentals of Data Visualization
book, for guidance alternatives to the Venn diagram. Neither book
discusses Venn diagrams, but they both provide valuable insight into
creating bar plots. Interestingly, the R for Data Science book does use
Venn diagrams to illustrate the differences between
full_join, so that provides some evidence that Venn diagrams are
useful for conceptualizing ideas about data.)
The UpSet plot
has gained some popularity since 2016. I’ve made a handful of UpSet
upsetR. While it is
possible to modify the aesthetics of an UpSet plot, it is very
challenging to manipulate the size, shape, colors, and font to the
degree that is needed to combine many plots into one multi-panel figure
for a manuscript.
In this repository, I turn 12 Venn diagrams from published research
articles to four bar plots made with
cowplot. I think
these bar plots can be made more simplicity, flexibility, reliability,
and reproducibility than a Venn diagram.
Let me know what you think!
Example 1 is inspired by “Cognitive specialization for learning faces is associated with shifts in the brain transcriptome of a social wasp”" by Berens et al. 2017.
Example 2 is from “Transcriptomics of an extended phenotype: parasite manipulation of wasp social behaviour shifts expression of caste-related genes”" by Geffre et al.
Example 3 is from “Sex-biased transcriptomic response of the reproductive axis to stress”" by Calisi et al 2017.
Example 4 is from “Early life stress alters transcriptomic patterning across reward circuitry in male and female mice”" by Peña et al. 2017.
So, that’s my comparison of four Venn diagrams and four possible bar plot alternatives. What do you think? Do you see the advantages? What other alternatives do you suggest? Comments welcome!