Package: TweedSpot 0.99.1

TweedSpot: Self-Adaptive Spatial Tweedie Models for Spatially Variable Gene Detection

TweedSpot detects spatially variable genes (SVGs) in spatially resolved transcriptomics data using self-adaptive Tweedie generalized additive models with nonparametric spatial smoothing. The method models raw counts directly, avoiding the bias introduced by log-transformation, adapts a gene-specific Tweedie index parameter to the mean-variance relationship of each gene, and offers an optional two-part formulation to accommodate excess zeros. Inputs and outputs use the SpatialExperiment infrastructure for interoperability with the Bioconductor ecosystem.

Authors:Nalin Arora [aut, cre], Himel Mallick [aut], Vivian Myers [aut]