Comprehensive Analysis Skill

Generates publication-quality visualizations and comprehensive analysis of Oxford Nanopore sequencing experiments with consistent end-reason color coding and intelligent data sampling.

Features

• KDE-based distributions: Smooth kernel density estimation with semi-transparent end-reason overlays
• Consistent color scheme: End-reason colors maintained across all 9 figures
• Data sampling: Default 50K reads (stratified by end_reason) for quick analysis
• Runtime estimation: Estimates full analysis time based on sampled execution
• Publication-ready: Up to 600 DPI output with proper annotations
• Interactive dashboard: HTML dashboard with modal viewing

Quick Start

# Quick analysis (default - samples 50K reads)
python comprehensive_analysis.py /path/to/experiment -o output/

# Full analysis (all reads - for publication)
python comprehensive_analysis.py /path/to/experiment -o output/ --full

# Custom sample size
python comprehensive_analysis.py /path/to/experiment -o output/ --sample-size 100000

# From sequencing summary directly
python comprehensive_analysis.py --summary /path/to/sequencing_summary.txt -o output/

# With options
python comprehensive_analysis.py /path/to/experiment \
    --output output/ \
    --dpi 600 \
    --title "My Experiment" \
    --open-browser \
    --full

Command Line Options

Data Sampling

By default, the skill samples up to 50,000 reads for quick visualization:

• Stratified sampling: Maintains end_reason proportions
• Statistics from full data: All statistics always computed from full dataset
• Runtime estimation: Estimates time for full analysis based on sampled execution
• Annotated figures: Sampled figures clearly marked with read counts
• Dashboard banner: Orange banner shows sampling status and estimated full runtime

Generated Figures (9 total)

End Reason Color Scheme

Output Structure

output/
├── figures/
│   ├── 01_overview_summary.png
│   ├── 02_length_by_end_reason.png
│   ├── 03_quality_by_end_reason.png
│   ├── 04_quality_length_correlation.png
│   ├── 05_2d_density_by_end_reason.png
│   ├── 06_temporal_evolution.png
│   ├── 07_hourly_evolution.png
│   ├── 08_channel_analysis.png
│   └── 09_publication_summary.png
├── statistics.json
└── dashboard.html

Statistics JSON

{
  "total_reads": 45136865,
  "sampled_reads": 50014,
  "is_sampled": true,
  "total_bases": 30479300000,
  "mean_length": 675.2,
  "median_length": 450.0,
  "n50": 675,
  "mean_qscore": 11.5,
  "median_qscore": 12.1,
  "pass_reads": 45136865,
  "fail_reads": 0,
  "pass_rate": 100.0,
  "end_reasons": {
    "data_service_unblock_mux_change": 41251754,
    "signal_positive": 3734572,
    "unblock_mux_change": 103126,
    "signal_negative": 44003,
    "mux_change": 2892
  }
}

Integration with ont-experiments

Run through ont-experiments for provenance tracking:

ont_experiments.py run comprehensive_analysis exp-abc123 \
    --output results/ \
    --full \
    --dpi 600

Example Output

Loading: /path/to/sequencing_summary.txt
  Loaded 45,136,865 reads

  [SAMPLING] Dataset has 45,136,865 reads > 50,000 threshold
  [SAMPLING] Using stratified sampling by end_reason
  [SAMPLING] Using 50,014 sampled reads for visualization
  [SAMPLING] NOTE: Run with --full flag to use all 45,136,865 reads

=== Statistics (from full dataset) ===
Reads: 45,136,865
Bases: 30479.3Mb
N50: 675 bp
Mean Q: 11.5

=== Generating Figures ===
  Generated: 01_overview_summary.png
  ...
  Generated: 09_publication_summary.png

  Figure generation time: 32.4s
  Estimated time for full 45,136,865 reads: ~9.8 hours

=== Analysis Complete ===