1 折线图
Code
f4h <-read_excel("data/01source.xlsx", sheet = "Fig.4", range = "A20:P23")
colnames(f4h) <-
c("Days", rep(
c("PBS", "SLAMF6+ PD-1+ CD8+ cells", "SLAMF6- PD-1+ CD8+ cells"),
each = 5
))
f4h |> pivot_longer(cols = -1,
names_to = "method",
values_to = "volume",) |>
dplyr::mutate(id=factor(rep(1:15,3)),
Days = factor(Days, levels = c("-1", "7", "14")),
method = factor(
method,
levels = c("PBS", "SLAMF6+ PD-1+ CD8+ cells",
"SLAMF6- PD-1+ CD8+ cells")
)) -> f4h
f4h
#> # A tibble: 45 × 4
#> Days method volume id
#> <fct> <fct> <dbl> <fct>
#> 1 -1 PBS 19.2 1
#> 2 -1 PBS 27.0 2
#> 3 -1 PBS 23.4 3
#> 4 -1 PBS 27.9 4
#> 5 -1 PBS 29.7 5
#> 6 -1 SLAMF6+ PD-1+ CD8+ cells 31.9 6
#> 7 -1 SLAMF6+ PD-1+ CD8+ cells 34.5 7
#> 8 -1 SLAMF6+ PD-1+ CD8+ cells 44.4 8
#> 9 -1 SLAMF6+ PD-1+ CD8+ cells 29.1 9
#> 10 -1 SLAMF6+ PD-1+ CD8+ cells 41.4 10
#> # ℹ 35 more rows统计p值
Code
# 重复测量方差设计
library(nlme)
model <- lme(volume ~ method*Days, random = ~1|id/Days, data = f4h)
summary(model)
#> Linear mixed-effects model fit by REML
#> Data: f4h
#> AIC BIC logLik
#> 304.5845 323.5868 -140.2923
#>
#> Random effects:
#> Formula: ~1 | id
#> (Intercept)
#> StdDev: 4.346709
#>
#> Formula: ~1 | Days %in% id
#> (Intercept) Residual
#> StdDev: 8.366812 3.060766
#>
#> Fixed effects: volume ~ method * Days
#> Value Std.Error DF t-value p-value
#> (Intercept) 25.43718 4.433186 24 5.737900 0.0000
#> methodSLAMF6+ PD-1+ CD8+ cells 10.81993 6.269472 12 1.725813 0.1100
#> methodSLAMF6- PD-1+ CD8+ cells 4.63341 6.269472 12 0.739043 0.4741
#> Days7 9.55505 5.634601 24 1.695781 0.1029
#> Days14 22.41864 5.634601 24 3.978745 0.0006
#> methodSLAMF6+ PD-1+ CD8+ cells:Days7 -37.58056 7.968529 24 -4.716123 0.0001
#> methodSLAMF6- PD-1+ CD8+ cells:Days7 -8.90525 7.968529 24 -1.117552 0.2748
#> methodSLAMF6+ PD-1+ CD8+ cells:Days14 -55.79891 7.968529 24 -7.002410 0.0000
#> methodSLAMF6- PD-1+ CD8+ cells:Days14 -14.68420 7.968529 24 -1.842774 0.0777
#> Correlation:
#> (Intr) mtSLAMF6+PD-1+CD8+c
#> methodSLAMF6+ PD-1+ CD8+ cells -0.707
#> methodSLAMF6- PD-1+ CD8+ cells -0.707 0.500
#> Days7 -0.636 0.449
#> Days14 -0.636 0.449
#> methodSLAMF6+ PD-1+ CD8+ cells:Days7 0.449 -0.636
#> methodSLAMF6- PD-1+ CD8+ cells:Days7 0.449 -0.318
#> methodSLAMF6+ PD-1+ CD8+ cells:Days14 0.449 -0.636
#> methodSLAMF6- PD-1+ CD8+ cells:Days14 0.449 -0.318
#> mtSLAMF6-PD-1+CD8+c Days7 Days14
#> methodSLAMF6+ PD-1+ CD8+ cells
#> methodSLAMF6- PD-1+ CD8+ cells
#> Days7 0.449
#> Days14 0.449 0.500
#> methodSLAMF6+ PD-1+ CD8+ cells:Days7 -0.318 -0.707 -0.354
#> methodSLAMF6- PD-1+ CD8+ cells:Days7 -0.636 -0.707 -0.354
#> methodSLAMF6+ PD-1+ CD8+ cells:Days14 -0.318 -0.354 -0.707
#> methodSLAMF6- PD-1+ CD8+ cells:Days14 -0.636 -0.354 -0.707
#> mSLAMF6+PD-1+CD8+c:D7
#> methodSLAMF6+ PD-1+ CD8+ cells
#> methodSLAMF6- PD-1+ CD8+ cells
#> Days7
#> Days14
#> methodSLAMF6+ PD-1+ CD8+ cells:Days7
#> methodSLAMF6- PD-1+ CD8+ cells:Days7 0.500
#> methodSLAMF6+ PD-1+ CD8+ cells:Days14 0.500
#> methodSLAMF6- PD-1+ CD8+ cells:Days14 0.250
#> mSLAMF6-PD-1+CD8+c:D7
#> methodSLAMF6+ PD-1+ CD8+ cells
#> methodSLAMF6- PD-1+ CD8+ cells
#> Days7
#> Days14
#> methodSLAMF6+ PD-1+ CD8+ cells:Days7
#> methodSLAMF6- PD-1+ CD8+ cells:Days7
#> methodSLAMF6+ PD-1+ CD8+ cells:Days14 0.250
#> methodSLAMF6- PD-1+ CD8+ cells:Days14 0.500
#> mSLAMF6+PD-1+CD8+c:D1
#> methodSLAMF6+ PD-1+ CD8+ cells
#> methodSLAMF6- PD-1+ CD8+ cells
#> Days7
#> Days14
#> methodSLAMF6+ PD-1+ CD8+ cells:Days7
#> methodSLAMF6- PD-1+ CD8+ cells:Days7
#> methodSLAMF6+ PD-1+ CD8+ cells:Days14
#> methodSLAMF6- PD-1+ CD8+ cells:Days14 0.500
#>
#> Standardized Within-Group Residuals:
#> Min Q1 Med Q3 Max
#> -0.62401382 -0.15515176 -0.03440848 0.17260010 0.72287328
#>
#> Number of Observations: 45
#> Number of Groups:
#> id Days %in% id
#> 15 45
df_aov <- anova(model)
df_aov
#> numDF denDF F-value p-value
#> (Intercept) 1 24 263.95442 <.0001
#> method 2 12 13.12729 0.0010
#> Days 2 24 1.89320 0.1724
#> method:Days 4 24 13.74242 <.0001
#成对比较
library(emmeans)
method_means <- emmeans(model, ~method)
print(method_means)
#> method emmean SE df lower.CL upper.CL
#> PBS 36.1 3.01 14 29.64 42.6
#> SLAMF6+ PD-1+ CD8+ cells 15.8 3.01 12 9.23 22.4
#> SLAMF6- PD-1+ CD8+ cells 32.9 3.01 12 26.30 39.4
#>
#> Results are averaged over the levels of: Days
#> Degrees-of-freedom method: containment
#> Confidence level used: 0.95
method_comparisons <- pairs(method_means)
method_comparisons
#> contrast estimate SE df
#> PBS - (SLAMF6+ PD-1+ CD8+ cells) 20.31 4.26 12
#> PBS - (SLAMF6- PD-1+ CD8+ cells) 3.23 4.26 12
#> (SLAMF6+ PD-1+ CD8+ cells) - (SLAMF6- PD-1+ CD8+ cells) -17.08 4.26 12
#> t.ratio p.value
#> 4.768 0.0012
#> 0.758 0.7345
#> -4.009 0.0046
#>
#> Results are averaged over the levels of: Days
#> Degrees-of-freedom method: containment
#> P value adjustment: tukey method for comparing a family of 3 estimates2 ggplot2
Code
pvalue <- method_comparisons |> broom::tidy() |>
dplyr::select(contrast, adj.p.value) |>
dplyr::mutate(padj=round(adj.p.value,digits = 4),
p_adj_sci=format(adj.p.value,scientific = T,digits = 4))
pvalue
#> # A tibble: 3 × 4
#> contrast adj.p.value padj p_adj_sci
#> <chr> <dbl> <dbl> <chr>
#> 1 PBS - (SLAMF6+ PD-1+ CD8+ cells) 0.00123 0.0012 1.230e-03
#> 2 PBS - (SLAMF6- PD-1+ CD8+ cells) 0.735 0.734 7.345e-01
#> 3 (SLAMF6+ PD-1+ CD8+ cells) - (SLAMF6- PD-1+ CD8+… 0.00456 0.0046 4.558e-03
p <- f4h |> summarise(
mean_volume = mean(volume),
sd = sd(volume),
.by = c(Days, method),) |>
ggplot(aes(
x = Days,
y = mean_volume,
color = method,
group = method,)) +
geom_errorbar(
aes(ymin = mean_volume - sd, ymax = mean_volume + sd),
position = position_dodge(0),
width = .2,
linewidth = 0.5,
) +
geom_line(position = position_dodge(0),
linewidth = 0.8) +
geom_point(
aes(shape = method),
position = position_dodge(0),
color = "black",
size = 2) +
theme_pubr() +
theme(legend.title = element_blank())
# 使用ggplot_build()获取图层数据
plot_data <- ggplot_build(p)
# 查看X轴的数值坐标
plot_data$data
#> [[1]]
#> ymin ymax x y colour group PANEL flipped_aes xmin
#> 1 21.2341315 29.640231 1 25.437181 #F8766D 1 1 FALSE 0.9666667
#> 2 29.8251093 42.689121 1 36.257115 #00BA38 2 1 FALSE 0.9666667
#> 3 23.2386460 36.902540 1 30.070593 #619CFF 3 1 FALSE 0.9666667
#> 4 20.3358912 49.648564 2 34.992228 #F8766D 1 1 FALSE 1.9666667
#> 5 4.8856152 11.577581 2 8.231598 #00BA38 2 1 FALSE 1.9666667
#> 6 22.3836755 39.057109 2 30.720392 #619CFF 3 1 FALSE 1.9666667
#> 7 28.7160676 66.995581 3 47.855824 #F8766D 1 1 FALSE 2.9666667
#> 8 0.5880804 5.165626 3 2.876853 #00BA38 2 1 FALSE 2.9666667
#> 9 27.2406302 48.369439 3 37.805035 #619CFF 3 1 FALSE 2.9666667
#> xmax order linewidth linetype width alpha
#> 1 1.033333 1 0.5 1 0.2 NA
#> 2 1.033333 2 0.5 1 0.2 NA
#> 3 1.033333 3 0.5 1 0.2 NA
#> 4 2.033333 1 0.5 1 0.2 NA
#> 5 2.033333 2 0.5 1 0.2 NA
#> 6 2.033333 3 0.5 1 0.2 NA
#> 7 3.033333 1 0.5 1 0.2 NA
#> 8 3.033333 2 0.5 1 0.2 NA
#> 9 3.033333 3 0.5 1 0.2 NA
#>
#> [[2]]
#> x y colour group PANEL flipped_aes order xmin xmax ymax
#> 1 1 25.437181 #F8766D 1 1 FALSE 1 1 1 25.437181
#> 2 2 34.992228 #F8766D 1 1 FALSE 1 2 2 34.992228
#> 3 3 47.855824 #F8766D 1 1 FALSE 1 3 3 47.855824
#> 4 1 36.257115 #00BA38 2 1 FALSE 2 1 1 36.257115
#> 5 2 8.231598 #00BA38 2 1 FALSE 2 2 2 8.231598
#> 6 3 2.876853 #00BA38 2 1 FALSE 2 3 3 2.876853
#> 7 1 30.070593 #619CFF 3 1 FALSE 3 1 1 30.070593
#> 8 2 30.720392 #619CFF 3 1 FALSE 3 2 2 30.720392
#> 9 3 37.805035 #619CFF 3 1 FALSE 3 3 3 37.805035
#> linewidth linetype alpha
#> 1 0.8 1 NA
#> 2 0.8 1 NA
#> 3 0.8 1 NA
#> 4 0.8 1 NA
#> 5 0.8 1 NA
#> 6 0.8 1 NA
#> 7 0.8 1 NA
#> 8 0.8 1 NA
#> 9 0.8 1 NA
#>
#> [[3]]
#> shape x y group PANEL order xmin xmax ymax colour fill size
#> 1 16 1 25.437181 1 1 1 1 1 25.437181 black NA 2
#> 2 17 1 36.257115 2 1 2 1 1 36.257115 black NA 2
#> 3 15 1 30.070593 3 1 3 1 1 30.070593 black NA 2
#> 4 16 2 34.992228 1 1 1 2 2 34.992228 black NA 2
#> 5 17 2 8.231598 2 1 2 2 2 8.231598 black NA 2
#> 6 15 2 30.720392 3 1 3 2 2 30.720392 black NA 2
#> 7 16 3 47.855824 1 1 1 3 3 47.855824 black NA 2
#> 8 17 3 2.876853 2 1 2 3 3 2.876853 black NA 2
#> 9 15 3 37.805035 3 1 3 3 3 37.805035 black NA 2
#> alpha stroke
#> 1 NA 0.5454545
#> 2 NA 0.5454545
#> 3 NA 0.5454545
#> 4 NA 0.5454545
#> 5 NA 0.5454545
#> 6 NA 0.5454545
#> 7 NA 0.5454545
#> 8 NA 0.5454545
#> 9 NA 0.5454545
p + annotate(geom = "segment",
x = 3.4,
y = 47.855824,
yend = 2.876853,) + annotate(
geom = "text",
x = 3.45,
y = (47.855824 + 2.876853) / 2,
angle = 270,
label = pvalue$padj[1],
) +
annotate(geom = "segment",
x = 3.15,
y = 47.855824,
yend = 37.805035,) + annotate(
geom = "text",
x = 3.2,
y = (47.855824 + 37.805035) / 2,
angle = 270,
label = pvalue$padj[2],
) +
annotate(geom = "segment",
x = 3.3,
y = 2.876853,
yend = 37.805035,) + annotate(
geom = "text",
x = 3.35,
y = (2.876853 + 37.805035) / 2,
angle = 270,
label = pvalue$padj[3],
)
Code
#y=c(47.855824,37.805035,2.876853,),
#yend = c(2.876853,47.855824,37.805035,),3 ggpubr
Code
Code
# Add p-values onto the line plots
# Remove brackets using linetype = "blank"
stat.test <- stat.test %>%
add_xy_position(fun = "mean_sd", x = "dose")
lp + stat_pvalue_manual(stat.test,
label = "p.adj.signif",
tip.length = 0,
linetype = "blank")