sc_analysis

Two focused pipelines for single-cell copy-number analysis:

Pipeline	Purpose	Entry point
post-scAbsolute QC	Classify cells, create auditable QC reports, and summarize dropout/read distributions	./run_scabsolute_qc.sh
post-scUnique visualization	Plot the evolutionary tree with final copy number and summarize private (freq == 1) events per cell	./run_post_scunique.sh

Historical analyses, old plotting scripts, and previous sample manifests are kept under archive/. They are preserved for reference but are not part of either maintained pipeline.

Repository layout

sc_analysis/
├── run_scabsolute_qc.sh
├── run_post_scunique.sh
├── environment.yml
├── pipelines/
│   ├── scabsolute-qc/
│   │   ├── run.sh
│   │   ├── R/
│   │   ├── scripts/
│   │   ├── config/
│   │   └── examples/
│   └── post-scunique/
│       ├── run.sh
│       └── scripts/
└── archive/

Installation

Create the shared Conda environment:

conda env create -f environment.yml
conda activate sc_analysis

Set RSCRIPT_BIN if Rscript is not on PATH:

export RSCRIPT_BIN=/path/to/Rscript

Pipeline 1: post-scAbsolute QC

This pipeline consumes scAbsolute QDNAseqCopyNumbers RDS objects. It performs cell-cycle classification, RPC filtering, MAPD/Gini/alpha QC, Borderline classification, manual-review profile reports, dropout summaries, and cross-sample reporting.

./run_scabsolute_qc.sh \
    path/to/samples.csv \
    path/to/project_data \
    500

The project data directory must contain:

project_data/
└── scAboslute-obj/
    └── SLX-<sample>_<bin_size>.rds

Results are written to:

• project_data/analysis_per_sample/ - cell classifications, QC tables, and per-sample PDFs.
• project_data/results_<project_name>/ - combined tables and cross-sample figures.

See pipelines/scabsolute-qc/README.md for arguments, QC order, thresholds, and outputs.

Pipeline 2: post-scUnique visualization

This pipeline consumes one completed scUnique sample directory and produces a tree-aligned final-copy-number heatmap plus the distribution of private events. Here, freq == 1 means an event locus was observed in exactly one cell in the analyzed scUnique cohort; a cell may carry several such private events.

./run_post_scunique.sh \
    path/to/scunique/results/SLX-27548_500 \
    path/to/output

Required files:

• <prefix>.finalCN.RDS
• <prefix>.tree.RDS
• <prefix>.df_pass_post.RDS

See pipelines/post-scunique/README.md for all generated figures and tables.

Sample manifest

The scAbsolute pipeline requires a CSV with a sample column. Optional metadata columns are used for grouped summaries:

sample,Cell line,feature1,feature2
27548,Example line,Group A,Condition A

Start from pipelines/scabsolute-qc/examples/sample_manifest.csv.