Pan-human Azimuth¶
Unified and scalable organism-wide cell annotation for single-cell and spatial transcriptomics
Important
Using Python? Start here.
Using R? Start here.
Out now on bioRxiv!
The Azimuth project represents a series of computational tools for reference-mapping of single-cell data. As a powerful complement to manual annotation and exploration workflows, reference-mapping pipelines aim to utilize existing knowledge to help automate the annotation and interpretation of new datasets.
We are excited to present Pan-human Azimuth, a neural network classifier that annotates human single-cell and single-nucleus RNA-sequencing experiments—across tissues and technologies—into a consistent and interpretable hierarchical cell typology. The model has been trained on data from 23 different tissues, encompassing 381 different cell types at the highest resolution, organized into a unified hierarchical cell typology based on an adapted version of the DISCO reference.
Pan-human Azimuth is made available through an open-source Python package (panhumanpy) and an R interface (AzimuthAPI), allowing users to annotate their datasets directly. The development of Pan-human Azimuth is led by the New York Genome Center Mapping Component as part of the NIH Human Biomolecular Atlas Project (HuBMAP).
A preprint describing the full methodology is now available on bioRxiv.