ResolVI: addressing noise and bias in spatial transcriptomics.

ResolVI improves spatial transcriptomics analysis by probabilistically correcting molecule misassignments, batch effects, and nuisance factors, enhancing cell state discrimination and spatial expression detection.

• Why it matters: Accurate assignment of molecules to cells is crucial for understanding tissue function and cell communication, but current methods often misassign molecules, limiting discovery of new cell states and subtle expression changes.
• What they did: The authors developed resolVI, a probabilistic model that operates after segmentation algorithms to correct molecule misassignments and other technical biases, applied across multiple datasets.
• The result: ResolVI significantly enhances the ability to distinguish cell states and detect subtle spatial expression differences, enabling more precise and integrated tissue analysis, and is openly available in scvi-tools.