.. Compocyte documentation master file ===================================== Compocyte ===================================== .. toctree:: :maxdepth: 2 :hidden: :caption: Tutorials tutorials/getting_started tutorials/01_training_PBMC_classifier tutorials/02_modifying_classifiers .. toctree:: :maxdepth: 2 :hidden: api/modules :toctree: generated :nosignatures: :caption: Important classes and methods Compocyte.core.hierarchical_classifier Compocyte.pretrained.pbmc_pretrained Compocyte.pretrained.til_pretrained .. toctree:: :maxdepth: 1 :hidden: :caption: Resources resources/troubleshooting resources/citation GitHub Repository Introduction ================== Compocyte is a composite classifier for modular hierarchical cell type annotation of single cell data. Using Compocyte you can build different hierarchical classifier architectures following a local classifier per parent node approach. Local classifiers are built around pytorch, sklearn or CatBoost. Local classifiers can be individually modified to account for alterations in classification taxonomies or selectively improve specific annotations in human-in-the-loop approaches. While compocyte has been primarily developed for single cell RNA sequencing data it can also be used with other single cell data compatible with the AnnData and scanpy packages. Quick links ================== .. grid:: 1 2 2 2 :gutter: 2 .. grid-item-card:: Getting started :link: tutorials/getting_started :link-type: doc Walk through practical examples and learn best practices. .. grid-item-card:: Tutorials :link: tutorials/index :link-type: doc Walk through practical examples and learn best practices. .. grid-item-card:: API Reference :link: api/modules :link-type: doc Complete documentation of all Compocyte modules and functions. .. grid-item-card:: Troubleshooting :link: resources/troubleshooting :link-type: doc Solutions to common issues and frequently asked questions. .. grid-item-card:: GitHub :link: https://github.com/WALL-E-Lab/Compocyte View source code, report issues, and contribute. Indices and tables ================== * :ref:`genindex` * :ref:`modindex` * :ref:`search`