feat: ci, web output, reorganization#22
Conversation
R script for parsing human interactome based on experimental evidence values R script for taking PANTHER pathways and parsing into STRING ID edge list
Now can access a list of TFs and Receptors to process a PANTHER pathway into sources and targets.
Used the function to generate Source and Target sets for 14 PANTHER pathways
…ss the data to make the combined networks (still in progress)
|
There is a lot here - I don't have the technical chops to do a full code review without talking through the PR with @tristan-f-r and @AMINOexe. I made a few comments, and look forward to talking more. One question for @agitter is whether we should perform this merge as a first-attempt at the repo (since the repo is so new), knowing that not all details have been thought out yet. It would help us get a sense of how to run multiple datasets, and would be helpful to have this merged when we think about file storage solutions. |
|
The synthetic pathways dataset and Gabes ResponseNet dataset still need a review over the code in general. I think we should open PRs again to review those once this is merged rather than trying to review those in this PR. |
|
[Waited for CI to pass] I've removed the unreviewed datasets and am splitting it to a separate PR. |
(TLDR): This primarily integrates CI, which runs the
yeast-osmotic-stress,hiv, andresponsenetdatasets, displaying them on a little webpage.This supersedes #5 and #11 (and, consequently, #7 and #6) by integrating the RN data and the synthetic pathway data into this pull request (for testing the PRA workflow). Note: This PR contains scripts (including from #5 and #11) that process biological data and should be reviewed.
Organization
This closes #21, with the following organization:
configsis our SPRAS configs, split by data type (DMMM, PRA)datasetsis each category of datasets, and is mainly meant for heuristic organization (e.g.hivandyeast-osmotic-stressare still separate, but synthetic networks that come from the same database would be categorized in the samedatasetfolder. This isn't an all-encompassing standard and should be broken when necessary)sprasas a submodule (it is easier to make commits tospras-benchmarkingthan it is tospras)webfor web outputEventually, there will be an
analysisfolder for further analysis not yet supported by SPRAS, making SPRAS-benchmarking double-serve as a staging area for analysis.