#Microbiome-graph
Produces a network of given microbiome data
These programs were written for the BioSeq program at Tufts University, as part of Hannah's summer research project on the microbiomes of various probiotics.
###Step 1: Node and edge table production in Python
This python script uses an input CSV of aggregate counts for each genus of bacteria found in a set of samples.
Utilizing this CSV file, a dictionary is made: The key being a sample name, and for each key, a set of ten tuples are made, with the first value being the Genus, and the second value being a float of the percent abundance in the sample.
The edges table has the following: From Node, To Node, and Weight. The node table has the following: Node, Type (Sample vs OTU)
Iterating through this dictionary, two CSV files are written and exported: one for the edge table, and one for the node table.
###Step 2: Producing an aesthetically pleasing graph in R
Using the csv files produced in Python, and the iGraph package in R, a pdf of a weighted graph is produced. Colors are used to differentiate between a genus and a sample. Formatting may be required to adjust to the number of samples.
Requirements: installation of the igraph and extrafont packages
Currently working with the Gephi interface in order to have a "dynamic" representation of data. With the official microbiome run data, the scripts above were very successful!