Fludarabine MRD Analysis

This repository contains the source code for the CLL MRD growth kinetics projects.

Setup

1. Install Python dependencies:

pip install -r requirements.txt

2. Authenticate with Google Cloud (required for accessing Terra workspaces):

gcloud auth login

Repository layout

• notebooks/ – all Jupyter notebooks grouped by project (crc, fi, gcll, misc, and autogen).
  • gcll/active/ holds the working set of GCLL notebooks.
  • gcll/archive/ retains legacy notebook snapshots for reference.
• data/ – shared inputs arranged by study: gcll/ (cell counts, treatment logs, spreadsheets), crc/inputs/ & crc/cnv/, fi/, rnaqc/, and wbc_variance/.
• outputs/ – generated growth kinetics reports exported from notebooks
• external/ – third-party tools vendored into the repo (comut, signature_analyzer).
• Cell_Population/ plus helper scripts (helper.py, model_helper.py, plotly_helper.py) remain at the top level for backward compatibility with existing notebooks.

Scripts

• scripts/run_cell_population.py – fetches the Cell Population inputs (cluster CCFs, mutation CCFs, sif, build_tree_posteriors) for a patient from Terra workspace via dalmatian, then runs python run_cell_population.py. Example:

  scripts/run_cell_population.py CLL14-1056 broad-firecloud-ibmwatson/TAG_CLL_Clonal_Kinetic_UMI_PrAN --tree-number 4

• scripts/build_patient_report.py – renders the Plotly report from a JSON configuration. Point it at a config (single patient or list) with your data paths and options, then run:

  python scripts/build_patient_report.py --config data/patient_config.json

  Reports land in outputs/current/html_reports/.