fusilli-multiqc

Custom MultiQC modules for the FUSILLI fusion library pipeline.

This package is made to be run with the FUSILLI pipeline: https://github.com/user/fusilli. It provides metrics and visualizations specific to the pipeline's outputs.

• Detection Metrics: Detection efficiency, library coverage, and sensitivity analysis
• Diversity Metrics: Library diversity metrics
• Processing Metrics: Read and base retention and loss through processing steps
• Partner Detection: Partner detection metrics and coverage across samples

Installation

Through FUSILLI

The FUSILLI pipeline uses a conda environment to install multiqc and this module. You should not need to install this module separately if you are using FUSILLI.

Through pip

pip install fusilli-multiqc

Requirements

• Python >= 3.9
• MultiQC >= 1.25
• pandas >= 1.0
• numpy >= 1.0

Usage

After installation, the modules will be automatically discovered by MultiQC when you run:

multiqc <results_directory>

The modules will automatically detect and parse the following FUSILLI output files:

• fusion_qc_metrics.csv - Detection efficiency and coverage metrics
• sensitivity_metrics.csv - Sensitivity analysis data
• fusion_counts_summary.csv - Fusion count data for diversity analysis
• decay_metrics.csv - Read decay through preprocessing steps
• trim_metrics.csv, contam_metrics.csv, quality_metrics.csv - Per-filter read/base removal
• unmerged_qc_metrics.csv - Pre-merge (unmerged R1/R2) detection metrics
• partner_counts_summary.csv - Partner detection counts

Module Descriptions

Detection Metrics Module (fusilli_detection)

Visualizes:

• Detection Efficiency: Fraction of processed reads that matched fusion sequences
• Prefilter Efficiency: Fraction of reads containing partner domain 3' ends
• Matching Efficiency: Fraction of prefiltered reads that matched specific breakpoints
• Library Coverage: Variant, breakpoint, and partner coverage metrics
• Sensitivity Analysis: Sensitivity index vs expected detection fraction
• Unmerged Read QC: Detection counts recovered from pre-merge R1/R2 reads

Diversity Metrics Module (fusilli_diversity)

Visualizes:

• Shannon Diversity: Measures both richness and evenness
• Simpson Diversity: Measures dominance
• Evenness: Pielou's evenness index
• Top N Fractions: Distribution of counts across top variants
• Variant Count Distribution: Histogram of variant abundances

Preprocessing Metrics Module (fusilli_preprocessing)

Visualizes:

• Read Decay: Read counts at each preprocessing step (log scale)
• Retention Rates: Fraction of reads retained after each step
• Step Loss Breakdown: Cumulative loss of reads through preprocessing
• Filtering Removal: Fraction of reads removed by the contaminant, adapter-trim, and quality filters

Partner Detection Module (fusilli_partners)

Visualizes:

• Partner Detection Heatmap: Binary detection matrix of partners across samples
• Partner Coverage: Number of samples with detection per partner
• Partner End vs Linker: Comparison of detection methods

License

MIT License - see LICENSE file for details.