This gene browser offers a convenient way to interact with our SFO scRNA-seq dataset and visualize gene expression across different thirst conditions. Expression data from each condition are integrated into a single expression matrix using Harmonypy. Gene expression count is normalized to 10,000 reads, log transformed, and scaled. Dimensionality reduction with principal component analysis is performed on the scaled gene expression data. 20 principal components are used as input to clustering analysis. The shared nearest-neighbor graph is compiled and transcriptomic clusters are identified by optimizing the graph modularity function with the Leiden algorithm as implemented by the scanpy.pp.neighbors and scanpy.tl.leiden functions (resolution parameter of 2). The results of clustering are visualized in both a two-dimensional and interactive three-dimentional UMAP embedding.