README

*** Use main.py to interact with the program

Comparison of taxonomic trees induced by group of samples.

**** Algorithm to get the most different pairs of (groups of) patients:

- Choose a metadatum such as its values can induce a partition of the set of samples
(Or only select groups of samples by their names)

- Construct the complete graph of these trees where edges are weighted by a certain comparison function (Zhang-Shasha's algorithm, total ratio of nodes, ...)

- Apply Kruskal's algorithm on this graph to obtain the minimal covering tree

- Store distances in a matrix and select the pairs of groups de patients that are the most different (which distance is superior to the third quartile of the distances)
(Default distance is LCA-path-length difference)

**** Files:
- actions.py

- comparisonFunctions.py
Implementation of comparison functions.

- distanceFunctions.py
Implementation of distance functions.

- /files
Storing results.

- graph.py
Implementation of graph (of the trees).

- kruskal.py
Implementation of Kruskal's algorithm.

- main.py
Interface with the user.

- /meta
Contains metadata.

- misc.py
Useful functions.

- parsingInfo.py
Parses the data matrix.

- parsingMatrix.py
Parses the occurrence matrix.

- parsingTree.py
Parses the taxonomic tree.

- taxoTree.py
Implementation of the taxonomic tree.

- writeOnFiles.py
Self-explanatory.