How unique379r/strspy works
Project Type and Purpose
• This is an open‐source bioinformatics pipeline tool (STRspy) for short tandem repeat (STR) detection in long‐read sequencing data (ONT or PacBio).
• It provides a command‐line interface, uses Conda environments, calls external tools, and generates STR quantification, variant calls, and visualizations.
File Structure
• Top‐level directories:
• config – user input/configuration files
• customDB – STR reference databases (Fasta/Bed)
• scripts – main pipeline scripts (parallel and nested versions)
• setup – environment and tool‐config setup scripts
• utilities – post‐processing and plotting scripts
• demodata – example data and results
• figs – static figures (logo, benchmarking images)
• Key files:
• STRspy_run_v1.0.sh – wrapper launcher
• scripts/STRspy_v1.0.sh – core pipeline
• setup/STRspy_setup.sh & MakeToolConfig.sh – environment and ToolsConfig
• External tools invoked (configured in config/ToolsConfig.txt): minimap2, samtools, bedtools, xatlas, GNU parallel.
README for Architectural Insights
• Inputs: FASTQ or BAM, STR Fasta/Bed, genome reference, normalization cutoff.
• Workflow steps: alignment → coverage estimation → allele counting & normalization → SNP/indel calling → stutter analysis → reporting.
• Outputs: allele counts, normalized counts, mapping summary, stutter plots, heatmaps.
Main Components
a. Input Layer
• User CLI (STRspy_run_v1.0.sh)
• config/InputConfig.txt and config/ToolsConfig.txt
• Raw sequencing data directory (FASTQ/BAM)
b. Setup Layer
• Conda environment (STRspy_env)
• setup scripts (environment.yml, STRspy_setup.sh, MakeToolConfig.sh)
c. Reference Data
• customDB/Y-str-prep (STR Fasta/Bed)
d. Pipeline Engine
• scripts/STRspy_v1.0.sh (core workflow)
• GNU parallel (optional parallel execution)
e. External Bioinformatics Tools
• minimap2 for alignment
• samtools and bedtools for coverage and mapping summary
• xatlas for variant calling
f. Post‐Processing Utilities
• utilities/*.sh and *.py/.R for benchmarking, stutter analysis, plotting
g. Output Layer
• OUTPUT_DIR: raw allele counts, normalized counts, variant reports, plots
• demodata/test_results for comparison
Relationships and Data Flow
• From Pipeline Engine to Reference Data (STR DB) for region definitions.
• From Pipeline Engine to Post‐Processing Utilities, then to Output Layer.
Architectural Patterns and Design Principles
• Modular pipeline: separation of setup, core workflow, utilities.
• Configuration‐driven: all paths and tool executables defined in config files.
• Extensible reference database (customDB) separate from code.
• Optional parallelization layer (GNU parallel) for scalability.
Technologies and Frameworks
• Conda for environment management
• Bash scripts orchestrating workflow
• Python and R for plotting and analysis (utilities)
• Third‐party genomics tools: minimap2, samtools, bedtools, xatlas, GNU parallel
Essential Elements
• Component names and roles
• Version or tool‐name annotations near external tools