MDS Plotting function - Githubissues

Shians / NanoMethViz

Apache License 2.0

21 stars 2 forks source link

MDS Plotting function #9

Open Shians opened 2 years ago

Shians commented 2 years ago

Currently, there is only support to create the counts matrix to be used for MDS plotting. It would be nice to have a complete function that can take a methy object and directly produces an MDS plot.

Technical concerns are:

Converting to counts from methylation tables time consuming, re-doing work each time we plot is not ideal.
Decisions to be made on counting at a per-site level vs per-gene level and the subsequent feature selection.
- Most likely his limma's MDS, which requires a number of top features. The default probably not appropriate for site level counts.

mpodio81 commented 3 days ago

Hi, Im having an issue with the MDS plotting... The labels change if I use: plot_mds(lmr) + ggtitle("MDS Plot")

or this:

plot_mds(lmr, groups = group, labels = group) + ggtitle("MDS Plot") + scale_colour_brewer(palette = "Set1")

And I dont know which is the right one....

Shians commented 3 days ago

Hi, Im having an issue with the MDS plotting... The labels change if I use: plot_mds(lmr) + ggtitle("MDS Plot")

or this:

plot_mds(lmr, groups = group, labels = group) + ggtitle("MDS Plot") + scale_colour_brewer(palette = "Set1")

And I dont know which is the right one....

@mpodio81, the top one uses the sample names as labels, while the bottom uses the group names. If you had more than one sample per group then the bottom one would have duplicated labels for all samples within the same group. For your case it's up to you whether you want to show group names or sample names.