StrainPhlan plots questions

Microbial community profiling StrainPhlAn
1

Hi,

I used R to make the plots for StrainPhlan outputs. In the plot, I found some control samples are all “-” in sequence. What does this mean?

image
image1030×458 51.8 KB

In the ordination plot, PCo1 explained 109% variance, then I checked eigenvalues and found outliers. What should I do for those outliers?

image
image582×517 29 KB

image

Thanks,
Ai

2

Hi @Ai_Zhang
For the first plot with the tree, you should notice that you are showing few positions of your alignment, that, in your case, belong to marker(s) not present in some of your strains. For a better (and full) visualisation of the tree + MSA, I would suggest you to use: Tree viewer - Online visualization of phylogenetic trees (newick) and alignments.
Regarding your PCA, is the boxplot the representation of all the eigenvalues?

3

Hi Aitor,

I use the tree viewer and it looks better.
Yes, the boxplot showed the eigenvalues that I got using Rcripts:
e.sir.pcoa ← cmdscale( data_Clostri_distmat, eig = T )
boxplot(eigenvals(e.sir.pcoa))

Thanks,
Ai

4

Hi @Ai_Zhang
I see, if that is the case, I still don’t get what is the problem to found the outliers in the eigenvalues distribution. The eigenvalues are related to the explained variance of each component, and, as you saw, PCo1 explains practically all the variance of your data (the top outlier)