Fabian Theis (Helmholtz Zentrum Munich)
To expand Scanpy’s core infrastructure and community platforms for stability, versatility, and sharing knowledge.
Single Cell ANalysis for PYthon (Scanpy) is a Python framework for preprocessing, analyzing, and visualizing single-cell data. It has proven to be a fast and popular interactive analysis environment as well as a base for other single-cell libraries. The team will improve Scanpy’s existing infrastructure and community management as well as extend Scanpy to provide effective analysis and visualization to a broader community, in particular with respect to spatial transcriptomics. Scanpy is the second most popular single-cell analysis software after Seurat. Its success is based on its comprehensiveness, speed, ease of learning, and usage of Python. In the past, the feature-driven development accumulated technical debt and potential instability, so this work will be invested into maintenance — for example, by increasing test coverage and building protections against performance regression. Additionally, in order for the Scanpy community to grow, documentation, community standards and social platforms will play a vital role. Spatial transcriptomics is quickly growing in impact and recently became commercially available. Scanpy will need both the appropriate abstractions to represent spatial information and reliable implementations of best-of-breed preprocessing, analysis, and visualization for this data.