Spatially variable gene detection

Unlike prior methods that leverage separate clustering algorithms (e.g., mclust, K-means) on learned representations, RIF seamlessly connects representation learning and SDI, which enables interpretation of the trained network to identify (Spatially variable genes) SVGs that play a significant role in distinguishing spatial domains.

• 10X Visium DLPFC
• 10X Visium HBRC
• Stereo-seq MOB