Backlog

In Progress

• working on conda install (stable-conda-env branch. JP built the environment.yml file CS and Vivan are trying to install it)
• running the DMD --> biomarker pipeline
  • make more complicated/better pdf generation

General things to change in the code

• add more information to the steps field of the chatlog in all files
  • implement code that will produce a python script to run the analysis run in BRAD (this is coming along fairly well for the running code module, but I'm not sure how to integrate enricher, scraping etc.)

Backlog of items for BRAD Code Base

• improve the user interface
• improve llm output parsing:
  • pydantic: ensures typing of llm output
  • json parser: can format llm output
  • implement difflib to parse llm output that is multiple choice
• work on data visualization
• put planner call into the planner module
• add images to BRAD retrieval
• how to get python script output to display both to the UI and appear in the log? I could just print it after

Backlog of code to write that will work with BRAD

• Hardwired Genome python codes (Joshua + Yuchen)
• Codes for dmd/dgc models + sensor selection are in the BRAD-Tools repository
• can we write some code to make it do basic machine learning problems?
• make Nicks code for system identification runnable with BRAD