Extended Data Fig. 10: Dataset quality assessment: Genes, UMI counts, and percentage of mitochondrial genes per cell split by cluster and by dataset.
From: Neuroinflammatory astrocyte subtypes in the mouse brain
t-SNE projections of astrocytes and violin plots of quality control metrics for AD models (5xFAD, Trem2-/-, and Trem2-/-/5xFAD) (a), a model of demyelination (experimental autoimmune encephalomyelitis, EAE) (b), and an acute injury/stab wound model (c) split by astrocyte cluster (top) and by dataset (bottom). Habib et al. (2020 – 5xFAD)34 data are in purple, Zhou et al. (2020 – 5xFAD, Trem2-/-, and Trem2-/-/5xFAD)35 7-month time point data are in red, Zhou et al. (2020 – 5xFAD) 15-month time point data are in yellow, Wheeler et al. (2020 – EAE)36 B6 sample data are in green, Wheeler et al. (2020 – EAE) TdTomatoGFAP-sorted sample data are in orange, Zamboni et al. (2020 – stab wound)37 data are in pink, and the current study data (acute inflammation using lipopolysaccharide, LPS) are in blue. Abbreviations: AD, Alzheimer’s disease; B6, C57BL/6 J mice; GFAP, TdTomatoGFAP mice; EAE, experimental autoimmune encephalomyelitis; LPS, lipopolysaccharide; Trem2-/-, Trem2 knockout mice; Trem2-/-/5X, Trem2 knockout 5xFAD mice; UMI, unique molecular identifier.
