Proposed Repo Structure #21
Description
This is a first stab at standardizing the file structure of this repo. Looks like we currently have a structure like:
case-study-1/
case-study-2/
synethic-pathways/
where each top-level folder as raw-data/, processed-data/, outputs/, src/, and a README.md. Let's assume that each directory above gets the data formatted to be SPRAS-ready (w/ standard interactome files and sources/targets/prizes). We plan to add ResponseNet (#5 ) and BTB (from a SPRAS fork) with a similar structure.
Dataset Benchmarking: we propose checking SPRAS out as a submodule in spras-benchmarking/ and using a top-level config/ directory to house config/ files will be the best way to do analysis. We may have multiple config files for different experiments (pathway reconstruction, phosphoproteomics studies, edge orienting, module detection, etc.).
Analysis Output: we propose to have a top-level analysis-output/ directory for ensembling/clustering/etc. multiple datasets, which will be specified by the config files.
Right now it seems like the case study / synthetic pathway directories also do downstream analysis; these features could be retained or moved to analysis-output/ - let's discuss.