Outline:
Jonathan from the Alm lab has a nice package that you can use to visualized information from 16S libraries. A link to his website is here:
https://bitbucket.org/yonatanf/survey/wiki/Home
This is some notes that I've made for running it on beagle.
First, you have to use the -X option to log in:
ssh -X username...
Next, I found you can still use it on the dedicated node (don't qlogin):
ssh compute-1-1
Change into the working directory:
cd /your/directory
Call his main program (although I think his home dir is only visible to people in the Alm lab- I'll have to check that):
/home/yonatanf/alm_lib/mipython.sh
Now, I do what I need to do. This specifically makes an ordered heatmap of Mystic Lake depths:
In [1]: file = 'eOTU_parallel/ML_Q_unique_final.mat.lin'
In [2]: counts, lins = SM.fromTxt(file, lin=True)
In [3]: counts = counts.remove_rows(['MSB','MEB','M3.2','M8.2','M22'])
In [4]: fracs = counts.normalize()
In [5]: D = fracs.dist_mat(axis='rows', metric='JSsqrt')
In [6]: order = ['M21', 'M20', 'M19', 'M17', 'M16', 'M15', 'M14', 'M13', 'M12', 'M11', 'M10', 'M9', 'M8', 'M7', 'M6', 'M5', 'M4', 'M3.1', 'M1.5', 'M0']
In [7]: Dorder = D.reindex(index=order, columns=order)
In [8]: Dorder.plot_dist_heatmap(sort_rows=False, sort_cols=False, plot_row_labels=True, row_label_width = 0.06)