How jaredl7/soursop works
Project type and purpose
• SOURSOP is an open-source Python library for analyzing molecular dynamics simulations of intrinsically disordered proteins. It is not a web or mobile app, but a domain-specific analysis toolkit built on top of the mdtraj library.
• Its primary purpose is to provide a modular API for loading trajectory data, computing structural properties (e.g., radii of gyration, center‐of‐mass distances, secondary structure), and extending functionality via plugins.
File structure
• Top-level:
• soursop/ → core Python package
• demo_examples/ → Jupyter notebooks and figures showcasing library usage
• docs/ → Sphinx documentation
• devtools/ → scripts and Conda recipes for development, CI, packaging
• soursop/tests/ → pytest‐based unit and integration tests
• configuration files for CI (.travis.yml, GitHub workflows), packaging (setup.py, setup.cfg, MANIFEST.in), code quality (.codecov.yml, .lgtm.yml)
• Within soursop/:
• __init__.py + soursop.py → top-level API entry points
• modules by function: sstrajectory.py, ssprotein.py, ssnmr.py, sspolymer.py, sspre.py, ssmutualinformation.py, etc.
• ssio.py, ssutils.py → I/O and general utility functions
• data/ → bundled test data and lookup tables
• plugins/ → plugin loader and example plugin (sparrow_plugin)
• _internal_data.py, configs.py → configuration and static data
• tests in soursop/tests/ verify each module’s behavior
README for additional insights
• Confirms mdtraj dependency and plugin extensibility
• Points to Sphinx docs for usage, examples, and developer guidelines (especially plugin development)
Elements
• Clear labels for: Core API, Functional Modules, Plugin System, External Library (mdtraj), Data Store, Testing & CI, Documentation
• User code → Core API → Functional Modules → mdtraj / Data
• Plugin loader ↔ Plugin modules
• CI/CD → Tests & Coverage → Reports
• Docs generator → Source code → HTML docs
• Blue rectangles for code modules
• Orange hexagon for external dependencies
Final advice
• Be thorough: include each top‐level directory as a bounded subsystem
• Data flow from user input to analysis output and reporting
• Do not overcomplicate: focus on separating distinct responsibilities (analysis modules, I/O, plugins, CI, docs)
• Use the README and module docstrings to annotate each component with its role