valdeanda / mebs

Protocol for finding informative protein families and then using them to score omic samples
17 stars 6 forks source link

empty visualizations #9

Open nvpatin opened 1 year ago

nvpatin commented 1 year ago

Hello,

Thanks for a very useful annotation program! I ran MEBS on a large data set of 550 MAGs. I got an output file that seems correct and complete (attached) but the .png files to visualize the results are empty. The heat map does have some names, a tree, and annotations, but the heat map itself is blank (attached).

Is my data set too large for MEBS to generate the visuals?

Thanks, Nastassia eCruises_TIMESERIES-MAGs-mebs-out tsv_comp_heatmap eCruises_TIMESERIES-MAGs-mebs-out.txt

eead-csic-compbio commented 1 year ago

My two cents, while @valdeanda or @acpooth reply, you might want to split your data in smaller sets and see if that works

valdeanda commented 1 year ago

Hi Nastassia Thank you for using MEBS, Yes, the problem is that to visualize that large number of MAGs you might want to try another type of visualization or change the figure size in the mebs_vis.py script To plot >100 MAGs I recommend you this heatmap script created by Cesar This is example input data , you can do columns as pathways and rows as genomes. You can even assign taxonomy or categories in the first column. Let me know if that works for you Val

nvpatin commented 1 year ago

Hi Val, Thanks for this advice. I used Cesar's heatmap.py script and it worked great! I can even use it for other types of heat maps, like MAG relative abundances across samples. Really helpful!! Nastassia