How aaramis/decaf works
Main Components
• CLI Interface: Represent the command-line interface (invoked via commands like “decaf process …”) that serves as the entry point for users.
• Core Library Modules: Showcase the key packages under the “decaf” directory, especially:
– Preprocessing: Handles input data transformation, filtering, and preparation.
– Utils: Provides shared utilities (data helper functions, common algorithms, etc.).
• Deep Learning / Classification Engine: Even if not explicitly separate, highlight the part of the core code that leverages deep learning to classify sequences and detect contaminants.
• File I/O & Format Support: Illustrate how FASTQ and FASTA files are received, processed, and output.
• Testing & Continuous Integration: Optionally include a component for unit tests (in the “tests” directory) and external workflows (as seen in the “.github/workflows” directory).
• Documentation & Configuration: Indicate that the project contains integrated documentation (using MkDocs in “docs” and “my-project/docs”) along with configuration files (pyproject.toml, setup.py, requirements.txt).
Relationships and Interactions
• The CLI routes the command to the appropriate backend modules: first into preprocessing where raw DNA sequence data (FASTQ/FASTA) gets ingested.
• Then the classification engine (core module possibly within preprocessing or a dedicated sub-component) processes the data, using deep learning models for filtering contaminant sequences.
• Utility modules assist in common tasks (such as logging, data transformation, or additional validations).
• Processed/Filtered output is then written back as cleaned files.
• Side modules handle tasks such as testing (auto-run via GitHub Actions) and documentation build/deployment.
Architectural Patterns and Design Principles
• Modular Design: Emphasize separation into distinct modules (CLI, preprocessing, utils) to allow extensibility for other genetic markers or taxa.
• Pipeline Structure: Data flows sequentially from input through preprocessing to deep learning classification and result output.
• Integration of DevOps: Use of CI/CD workflows for testing, linting, and documentation deployment ensures consistent quality.
• Clear Separation of Concerns helps in isolating core functionality from auxiliary services (environment configuration and documentation).
Relevant Technologies and Frameworks
• Python ecosystem (virtual environments, pip dependencies, black for code formatting).
• Command-Line Interface design (the “decaf” command in decaf-env/bin).
• Deep learning (while specific frameworks aren’t named, indicating “deep learning” is a core technology is important).
• MkDocs for documentation, as shown by mkdocs.yml and docs directories.
• GitHub Actions for continuous integration.