Search

Additional file 3: Figure S3. A comparison between the pattern of spot classification after CSDI through spot clustering and label transferring methods. The consistency between the results from these two methods, which is compatible with the histological information of tissue slices, supports the accuracy of the spot categorization.

Additional file 1: Figure S1. Improvement of the spot clustering through CSDI. The colors represent the results of unsupervised clustering performed either before (A) or after (B) application of CSDI. (A) The expected consistency of the pattern of spot clusters between consecutive slices is missed. (B) The consistency of clustering after consecutive slices data integration is improved. The patients IDs are marked above the pictures. The designations “before” and “after” refer to pictures made either before or after CSDI respectively.

Additional file 4: Figure S4. Improvement of the results from DEGs in spots assigned to neurons, before and after CSDI. Venn diagrams represent the intersection of spots assigned to neurons before (purple circle) and after (yellow circle) CSDI. In both situations, DEGs in neurons versus oligodendrocytes and astrocytes (if available) were applied to GO analysis and the results are shown in barplots.

Additional file 2: Figure S2. Improvement of the label transferring by CSDI. The annotation probability is shown as a scheme for three different cell types. Label transferring from snRNA-seq to ST consecutive slices is shown before and after data integration. Before Integration: the probability of spot annotations for neurons is not compatible with tissue histology. After integration: the probability of the presence of neurons increased in the GM of the cerebral cortex.