新的一周第一天，我们不整那些难的，来点小的绘图细节分析，如下图
图片.png
这个图包括两部分，左图的画法很简单
SpatialFeaturePlot(st_se, features = "Macrophage", pt.size.factor = 1,
crop = FALSE, alpha = c(0.1,1), min.cutoff = 0, max.cutoff = 1)
这是只是需要设置一下alpha参数，就可以把没有该细胞类型的位置给暴露出来。
关键是右边部分,其中st_pbs是包含预测的细胞类型分数矩阵的rds、Zone是空间聚类信息
zone_cell_avgs_pbs <- AverageExpression(st_pbs, assays = 'predictions', group.by = c('Zone', 'orig.ident'), slot = 'data') %>% as.data.frame()
zone_cell_avgs_pbs_melted <- zone_cell_avgs_pbs %>% melt()
zone_cell_avgs_pbs_melted$Celltype <- rep(rownames(zone_cell_avgs_pbs),12)
zone_cell_avgs_pbs_melted$Zone <- c(rep('Defect', 96), rep('IntactMuscle', 96), rep('Transition', 96))
zone_cell_avgs_pbs_melted$Tissue <- rep(c(rep('1197L', 24), rep('1203L', 24), rep('599L', 24), rep('600R', 24)), 3)
write.csv(zone_cell_avgs_pbs_melted, file = 'Data/PBS_Zone_Celltype_Averages.csv')
p <- zone_cell_avgs_pbs_melted %>%
filter(Celltype == 'Macrophage') %>%
ggplot(aes(x=Zone, y=value, fill=Zone)) +
stat_summary(fun.data="mean_se", geom="errorbar", width=0.5) +
stat_summary(fun="mean", geom="crossbar", aes(fill=Zone)) +
geom_point(aes(fill=Zone),size=4,shape=21, position = position_dodge(0.2)) +
labs(x="", y="Average Prediction Score") +
theme_classic() +
scale_x_discrete(limits = c("Defect", "Transition", "IntactMuscle")) +
scale_y_continuous(expand=c(0,0)) +
coord_cartesian(ylim=c(0,0.8)) +
stat_compare_means(comparisons =
list(c("Defect", "Transition"),
c("Transition", "IntactMuscle"),
c("Defect", "IntactMuscle")),
bracket.size = 1,
method = "t.test", paired = TRUE,
method.args = c(var.equal = FALSE),
label = "p.signif", size = 8) +
theme(axis.title.y = element_text(size = 20, face = "plain", color = "black")) +
theme(axis.text.y = element_text(size = 15, color = "black")) +
theme(axis.text.x = element_text(size = 20, colour = "black")) +
theme(legend.position="none") +
rotate_x_text(angle = 45)
print(p)
然后，大功告成了，
图片.png
生活很好，有你更好
