Fig. 4.

Application of SEDR on Stereo-seq dataset. A Laminar organization of a DAPI-stained mouse olfactory bulb. B Unsupervised clustering results from SEDR and nine competing methods on the olfactory bulb Stereo-seq data. C Visualization of individual identified clusters for selected methods (stLearn, SpaGCN, UTAG, DeepST, STAGATE, and SEDR). D Predicted olfactory bulb layers and corresponding marker genes. E Computational efficiency of ten methods tested on slice #151673 of the DLPFC set, and Stereo-seq data and Slide-seq data of mouse olfactory bulb. Left: run time of methods. Right: CPU memory usage of methods