Fig. 5: Chromatin and splicing patterns in AD.
a, Downsampling experiments. The distribution of the percentage of peaks that are significantly different between AD and control samples in the vicinity of genes targeted for splicing analysis is shown (Methods; n = 20). b, A peak that is highlighted within FMNL2 that is present in control astrocytes but not in AD astrocytes. Shading indicates peaks of interest. c, Downsampling experiment. The distribution of the percentage of exons showing significant differential inclusion per cell type in AD versus control is shown (Methods; n = 100). d, Percentage of novel reads found within control (n = 10) and AD (n = 9) datasets. e, Cell-type-resolved single-cell long reads for ZNF711. The top two tracks show AD excitatory neurons and control excitatory neurons, followed by AD oligodendrocytes and control oligodendrocytes. The bottom black track shows chromosome X: 85264898–85268508. f, Number of genes with one or more and two or more cell states detected in AD and control samples. Only the genes with testable exons were considered for cell-state detection. g, Volcano plot of state-specific exons across multiple cell types in AD and control groups (only exons with ten or more reads in two or more cell states were tested; n = 494 and 726). Exons with a P value of ≤0.05 and | LOR | of >1 are labeled in color, whereas the others are labeled in gray. A one-sided χ2 test followed by Benjamini–Yekutieli multiple testing correction was applied to evaluate the significance of splicing–cell state association. h, Density plot of the distribution of the maximum normalized-state ΔΨ per exon. Normalized-state ΔΨ = state ΔΨ /overall ΔΨ. i, Stacked bar plot showing the proportion of maximum normalized-state ΔΨ per exon split by value into three groups: <0.9, between 0.9 and 1 or ≥1. The ‘≥1’ group represents the fraction of disease-associated overall ΔΨ values, which can be seen in specific cell states by cell type. Each box plot shows the median (middle line), IQR (box) and adjacent values (whiskers extending to 1.5× the IQR). Dots represent outliers. A two-sided Wilcoxon rank-sum test was applied to all comparisons shown in a, c and d. FDR correction was applied to multiple comparisons, and corrected P values (<0.05) are shown.
