Single Cell Brain Atlas in Human Alzheimer’s Disease

One can visualize the single-cell metadata and gene expression overlaid onto a dimension reduction plot.

Gene Expression

Cell Information

One can visualize the single-cell gene expression grouped by a specified single-cell metadata. The resulting gene expression can be presented in a (i) bubbleplot where the colour represents the relative gene expression and the bubble size is the proportion of cells in the group expressing the gene, (ii) heatmap where the colour represent the average gene expression in each group, (iii) violinplot showing the distibution of non-zero gene expression in each group. Note: To speed up the violinplot, we only sampled a number of cells (nCells in the smallest group) from each group for plotting and this also ensures that there is the same number of cells within each group.

One can visualize the proportion / number of cells of a specified category in each group.

Cell Proportion

Dimension Reduction

One can visualize the single-cell metadata and gene expression of a particular cell type overlaid onto a dimension reduction plot.

Gene Expression

Cell Information

One can visualize the single-cell gene expression of a particular cell type grouped by a specified single-cell metadata. The resulting gene expression can be presented in a (i) bubbleplot where the colour represents the relative gene expression and the bubble size is the proportion of cells in the group expressing the gene, (ii) heatmap where the colour represent the average gene expression in each group, (iii) violinplot showing the distibution of non-zero gene expression in each group. Note: To speed up the violinplot, we only sampled a number of cells (nCells in the smallest group) from each group for plotting and this also ensures that there is the same number of cells within each group.

One can visualize the proportion / number of cells of a specified category in each group in a particular cell type.

Cell Proportion

Dimension Reduction

Differential expression and GSEA results for comparing the cell type of interest against other cell types.

Differential Expression

Gene Set Enrichment

Differential expression and GSEA results for comparing AD vs Control cells within each cell type. Cells are deemed AD / Control based on the library.

Differential Expression

Gene Set Enrichment

Differential expression and GSEA results for comparing AD vs Control cells within each cell type. Cells are deemed AD if its corresponding subcluster contains >80% cells from an AD-associated library and vice versa.

Differential Expression

Gene Set Enrichment

Differential expression and GSEA results for comparing the subcluster of interest against other subclusters within the same cell type.

Differential Expression

Gene Set Enrichment

Differential expression and GSEA results for comparing between pairs of subclusters within the same cell type. Here, we denote the DE/GSEA in a source_target format. Thus, a positive LFC/NES for m1_m2 indicates that a gene / pathway is upregulated in subcluster m2 as compared to subcluster m1.

Differential Expression

Gene Set Enrichment

List of GWAS genes used in this study, their Experimental Factor Ontology (EFO) categories [ Alzheimer's disease (EFO_0000249) / AD Biomarkers (EFO_0006514) / LOAD (EFO_1001870) / Neuropathologic (EFO_0006801) ] and associated PubMed IDs.
Differential expression of GWAS genes for comparing AD vs Control cells within each cell type. Cells are deemed AD / Control based on the library.

Differential Expression

Differential expression of GWAS genes for comparing AD vs Control cells within each cell type. Cells are deemed AD if its corresponding subcluster contains >80% cells from an AD-associated library and vice versa.

Differential Expression

Differential expression of GWAS genes for comparing the subcluster of interest against other subclusters within the same cell type.

Differential Expression

Differential expression of GWAS genes for comparing between pairs of subclusters within the same cell type. Here, we denote the DE in a source_target format. Thus, a positive LFC for m1_m2 indicates that a gene is upregulated in subcluster m2 as compared to subcluster m1.

Differential Expression

Gene Regulatory Network (GRN) scores for TFs predicted to regulate the transition between pairs of subclusters within the same cell type, calculated using the CellRouter algorithm (da Rocha et al. 2018). Here, we denote the transition in a source_target format. Thus, a positive GRN score for m1_m2 indicates that the upregulation of the TF is involved in the transition from m1 to m2. Note that for some transitions, no TFs are found after pruning the results. Furthermore, the CellRouter algorithm did not identify any TFs for endothelial cells and the algorithm was not ran on unidentified / hybrid cells.
List of TFs and their target genes for TFs predicted to regulate the transition between pairs of subclusters within the same cell type, calculated using the CellRouter algorithm (da Rocha et al. 2018). Here, we denote the transition in a source_target format. Note that for some transitions, no TFs are found after pruning the results. Furthermore, the CellRouter algorithm did not identify any TFs for endothelial cells and the algorithm was not ran on unidentified / hybrid cells.

[Citation to be inserted here] Currently Under Review.