Reorganize repo; add project setup#2
Conversation
- R scripts moved to r/ and numbered by workflow step (2_qc_check, 3_combine_batches, 4_outliers, 5_heatmap) - Python QC script and pose corner correction moved to python/ - QC config moved to config/QC_params.yaml - Exploratory script moved to notebooks/explore_features.py - pose_corner_correction.py: hardcoded paths replaced with --input_dir/--output_dir argparse args - Add pyproject.toml, uv.lock, .python-version for Python dependency management - Add renv.lock for R dependency management - README: full rewrite with module overview table and per-script documentation Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
| │ ├── 2_qc_check_cli.R same as above, CLI version for automation | ||
| │ ├── 3_combine_batches.R merge feature files + metadata into unified dataset | ||
| │ ├── 4_outliers.R outlier detection and QC figures (project-specific template) | ||
| │ ├── 5_heatmap.R phenotype correlation heatmaps |
There was a problem hiding this comment.
I think that the heatmap feature can be removed for now. The current version does not work and I'm not sure anyone is actually working on it. Jake may disagree though.
|
|
||
| ### 5 — Heatmap · `r/5_heatmap.R` | ||
|
|
||
| Generates phenotype correlation heatmaps. Under active development. |
There was a problem hiding this comment.
See my comment above. Is this actually something we want to include in the standard pipeline?
michberger
left a comment
michberger
left a comment
There was a problem hiding this comment.
I think this structure works. It is good to have all of the Nextflow/JABS related utilities we've developed in the same place. Is there much difference between the Python and R versions of step 2 (QC check)? If this is the only step that is in Python version, how would someone move onto the next step? Would they have to switch to R?
Also, as I indicated in comments, do we want to keep the Heatmap feature in the repository or assume that would fall under individual post-processing analysis. I'm not sure how useful a big correlation image really is.
2 participants
Huge structural changes!
Review README to get the overall view of the repo.
I intend to make it a mixture of R and python, building on Jake's pipeline and add several other utilities. Important changes: