spaceranger’s automated cluster number detection #3834
Description
What kind of feature would you like to request?
Additional function parameters / changed functionality / changed defaults?
Please describe your wishes
Dear scanpy Development Team,
I hope this email finds you well. First, I would like to express my sincere appreciation for developing and maintaining scanpy, which has become an indispensable tool for the single-cell and spatial transcriptomics community.
I am writing to suggest a feature that I believe would greatly benefit scanpy users working with spatial transcriptomics data. Currently, 10x Genomics' SpaceRanger includes an automatic clustering method that systematically evaluates different clustering parameters and selects the relatively optimal number of clusters. I would like to propose integrating a similar functionality into scanpy.
Rationale:
While manual parameter tuning gives users flexibility, an automated approach for initial clustering optimization would significantly streamline workflows, especially for users analyzing multiple datasets
SpaceRanger's automatic selection provides a reasonable starting point that users can then refine based on their specific biological knowledge
Although morphology-based classification remains one of the gold standards for spatial data validation, computational guidance for cluster number selection would help users make more informed initial decisions