Files with the suffix in .dependency.json are the inferred dependency graphs for the various databases. They are in JSON format and consist of lists of modules. Each module is a JSON object with the properties: 
    - id: an id identifying the module
    - clade: the ncbi id for the taxonomic category corresponding to the module or null
    - gains: a list of family ids gained in the module
    - losses: a list of family ids lost in the module
    - dependencies: a list of module ids that this module depends on
    - species: the ncbi id for the species or null

File with the suffix .tree.json are the trees used in the inference, after all adjustments described in the paper. They are in the same format as the .dependency.json files.

evaluate.py is a python script which can compute the log-likelihood of the graph or tree provided as input.
