Responsenet benchmarking - Gabe

.gitignore

@@ -1,3 +1,7 @@
+# Other
+*.sqlite
+*.DS_Store
+
 # Byte-compiled / optimized / DLL files
 __pycache__/
 *.py[cod]

responsenet-benchmarking/README.md

@@ -0,0 +1,34 @@
+# ResponseNet Benchmarking Overview
+
+This directory contains all the benchmarking data for the new SPRAS implementation of ResponseNet. 
+
+## Structure of Directory
+
+The sub-directory `/files` are `.json` outputs from the most recent version of the Yeger-Lotem's ResponseNetV.3, and `.txt` outputs from the SPRAS implementation of ResponseNet.
+
+The sub-directory `/output` contains various output files from the `prep_cyto.py` script. For each run of the script, two outputs are created. A nodefile and an edgefile.
+
+Finally, the sub-directory `/images` contains graph exports from cytoscape.
+
+### Original Paper:
+Yeger-Lotem E, Riva L, Su LJ, Gitler AD, Cashikar AG, King OD, Auluck PK, Geddie ML, Valastyan JS, Karger DR, Lindquist S, Fraenkel E. Bridging high-throughput genetic and transcriptional data reveals cellular responses to alpha-synuclein toxicity. Nat Genet. 2009 Mar;41(3):316-23. doi: 10.1038/ng.337. Epub 2009 Feb 22. PMID: 19234470; PMCID: PMC2733244.
+
+## Conda Environment
+
+The easiest way to install Python and required packages is with Anaconda.
+
+After installing Anaconda, you can run the following commands in this directory:
+```
+conda env create -f environment.yml
+conda activate responsenet_benchmarking
+```
+This will create a conda environment that is modified from the SPRAS environment, and then activate the environment. If you want to do it manually, install the latest versions of the following packages: `networkx`, `pybiomart`, and `pandas`.
+
+## Scripts
+There is only one script that needs to be run. This is assuming that files have been generated from ResponseNetV.3 and SPRAS. This script is: `prep_cyto.py`. You can call the script by running:
+```
+python prep_cyto.py --json [relative json filepath] --spras [relative spras filepath] --output [relative output filepath]
+```
+
+## Cytoscape Setup
+You should open the node and edge files into a cytoscape session. I recommend importing a network from an edge-file, and then adding the nodefile through nodetable import. The cytoscape session file in the directory showcases four networks.

responsenet-benchmarking/environment.yml

@@ -0,0 +1,30 @@
+name: responsenet_benchmarking
+channels:
+  - conda-forge
+dependencies:
+  - adjusttext=0.7.3.1
+  - bioconda::snakemake-minimal=7.19.1
+  - docker-py=5.0
+  - matplotlib=3.5
+  - pre-commit=2.20 # Only required for development
+  - pytest=7.1 # Only required for development
+  - python=3.8
+  - pip=22.1
+  - requests=2.28
+  - scikit-learn=1.2
+  - seaborn=0.12
+  - spython=0.2
+  # Needed for benchmarking/spras
+  - networkx=2.8
+  - pandas=1.4
+  - pybiomart=0.2.0
+  # Only required for GraphSpace
+  - commonmark=0.9
+  - docutils=0.18
+  - jinja2=3.1
+  - mock=4.0
+  - recommonmark=0.7
+  - sphinx=5.0
+  - pip:
+     - graphspace_python==1.3.1
+     - sphinx-rtd-theme==1.2.0

responsenet-benchmarking/files/Datasets/README.md

@@ -0,0 +1,15 @@
+## Information about Input Files
+
+ResponseNet has not been very clear about the data that they worked on. In this folder, all the files needed to replicate their work is found.
+
+#### File Breakdown
+The two files downloaded directly from ResponseNet are `ResponseNetNetwork.json` and `Muscle_Skeletal-Dec2018.tsv`. The Json file is an output from ResponseNet's sample output, and is what we used to compare to SPRAS.
+
+the `sources` and `targets` files are derived from the `ResponseNetNetwork.json`, we went through and scraped all the nodes in the json that were annotated as a source or a target node.
+
+The `Muscle_Skeletal-Dec2018.tsv` is the interactome that ResponseNet uses, they do provide a direct download on their site.
+
+#### Other information
+In order to download the files for yourself, you can do so at: http://netbio.bgu.ac.il/respnet
+
+You can directly download the interactome by selecting which one you are interested in using. In order to download their sample, you need to look for the link for the `sample output` and wait for ResponseNet to run. At the time of writing, ResponseNet will not allow you to directly download the source and target files, you must go to the cytoscape section of the software, and download the cytoscape `.json` file.