Biohacking: Comparing Multiple uBiome Samples
uBiome is providing a much needed service of sequencing our microbiome and trying to providing friendly user interfaces to review the results. Despite beautiful charts and comparison tools one of the limitations of uBiome’s web services is that you cannot see three or more samples at once. However, uBiome has provided some python tools through github to allow users to go further. Special thanks to Richard Sprague for providing so many tools on GitHub and a Microbiome Hackers Guide.
In order to merge three or more ubiome results we will need to do the following steps:
- download uBiome JSON file results
- install the uBiome python library
- install csvkit
- create a python script to merge the files
- view the results on the command line
Download uBiome JSON file results:
- Login to uBiome
- Click on Advanced
- Download Taxonomy JSON
- repeat for each date results are available for
Install the ubiome python library:
$ pip install ubiome
Install CSVKIT:
$ pip install csvkit
$ cat ubiome-multi-merge.py
Create a python script to merge the files:
#!/usr/bin/env python
from ubiome import *
f1 = UbiomeSample(“2016-11-21”)
f2 = UbiomeSample(“2017-04-01”)
f3 = UbiomeSample(“2018-01-29”)
m = UbiomeMultiSample(f1)
m.merge(f2)
m.merge(f3)
m.write(“merged-files.csv”)
Run the merge script and view the results:
$ ./ubiome-multi-merge.py
$ csvsort -c 3 -r merged-files.csv | csvlook
Once you have a merged CSV file you could of course use Google Sheets to view and sort the columns. My preference is for command line so I chose to use csvkit to compare my three years of microbiome results.