How pmenzel/ont-assembly-snake-mit works
Main System Components
a) Input Layer:
• ONT reads directory (fastq-ont/)
• Illumina reads directory (fastq-illumina/)
• Reference genomes directory (references/)
• Reference proteins directory (references-protein/)
b) Processing Layer (organize in parallel tracks):
• Read Filtering Track (Filtlong, Rasusa)
• Assembly Track (Flye, Raven, Miniasm, Unicycler, Canu)
• Polishing Track:
• Long-read polishing (Racon, Medaka)
• Short-read polishing (Pilon, Polypolish, POLCA)
• Reference-based polishing (Homopolish, Proovframe)
c) Output Layer:
• Final assemblies directory (assemblies/)
• Intermediate results
• Symlinks to final assemblies
Flow and Relationships
• Blue: Input/Output operations
• Green: Assembly processes
• Orange: Polishing processes
• Purple: Reference-based operations
• Gray: Configuration and environment
Implementation Details
• How sample names are parsed from directory structure
• How different tool combinations are determined from folder names
• Conditional execution paths based on available input data
Technical Elements
Components:
• GitHub Actions workflows (.github/workflows/)
• Conda environment specifications
• Snakefile as the central workflow controller