Une planche de graphiques
Spoiler alert
library(openxlsx)
library(ggplot2)
library(tidyr)
library(dplyr)
library(tibble)Importation et présentation des données
Abondance de 30 métabolites pour 38 observations réparties en 4 groupes
metabolite <- read.xlsx("metab.xlsx")
metabolite$groupe <- factor(metabolite$groupe,
levels=c("A1","B1","A2","B2"))
summary(metabolite)## M1 M2 M3 M4
## Min. :0.0005254 Min. :0.0003557 Min. :0.002925 Min. :0.001129
## 1st Qu.:0.0014447 1st Qu.:0.0006366 1st Qu.:0.005598 1st Qu.:0.001758
## Median :0.0019777 Median :0.0008990 Median :0.006657 Median :0.002077
## Mean :0.0023243 Mean :0.0012556 Mean :0.007003 Mean :0.002140
## 3rd Qu.:0.0029574 3rd Qu.:0.0018142 3rd Qu.:0.008238 3rd Qu.:0.002522
## Max. :0.0076899 Max. :0.0031358 Max. :0.013001 Max. :0.003504
## M5 M6 M7
## Min. :0.0009152 Min. :0.0006342 Min. :0.0000688
## 1st Qu.:0.0018662 1st Qu.:0.0010311 1st Qu.:0.0001098
## Median :0.0029078 Median :0.0012418 Median :0.0001535
## Mean :0.0030473 Mean :0.0014035 Mean :0.0001694
## 3rd Qu.:0.0040656 3rd Qu.:0.0016013 3rd Qu.:0.0002111
## Max. :0.0058096 Max. :0.0042977 Max. :0.0003652
## M8 M9 M10
## Min. :6.202e-05 Min. :0.0006151 Min. :0.0009801
## 1st Qu.:1.071e-04 1st Qu.:0.0029987 1st Qu.:0.0020337
## Median :1.406e-04 Median :0.0045993 Median :0.0023198
## Mean :1.507e-04 Mean :0.0064434 Mean :0.0025062
## 3rd Qu.:1.730e-04 3rd Qu.:0.0090120 3rd Qu.:0.0028410
## Max. :3.739e-04 Max. :0.0222578 Max. :0.0042863
## M11 M12 M13 M14
## Min. :0.0002614 Min. :0.002135 Min. :0.0002098 Min. :0.0001195
## 1st Qu.:0.0004246 1st Qu.:0.004089 1st Qu.:0.0003610 1st Qu.:0.0002138
## Median :0.0004975 Median :0.004573 Median :0.0004302 Median :0.0002469
## Mean :0.0005394 Mean :0.004587 Mean :0.0004959 Mean :0.0002764
## 3rd Qu.:0.0005895 3rd Qu.:0.005141 3rd Qu.:0.0005756 3rd Qu.:0.0003415
## Max. :0.0012367 Max. :0.006486 Max. :0.0010539 Max. :0.0005065
## M15 M16 M17
## Min. :0.0001815 Min. :2.291e-05 Min. :0.0002101
## 1st Qu.:0.0002695 1st Qu.:2.537e-04 1st Qu.:0.0004140
## Median :0.0003272 Median :4.641e-04 Median :0.0004816
## Mean :0.0003599 Mean :4.526e-04 Mean :0.0005101
## 3rd Qu.:0.0004250 3rd Qu.:5.702e-04 3rd Qu.:0.0005751
## Max. :0.0007758 Max. :1.197e-03 Max. :0.0009213
## M18 M19 M20 M21
## Min. :0.0001488 Min. :0.006853 Min. :0.0007242 Min. :0.0001917
## 1st Qu.:0.0002971 1st Qu.:0.011803 1st Qu.:0.0013885 1st Qu.:0.0004253
## Median :0.0003715 Median :0.016610 Median :0.0016649 Median :0.0005286
## Mean :0.0003806 Mean :0.019409 Mean :0.0018355 Mean :0.0005226
## 3rd Qu.:0.0004486 3rd Qu.:0.024679 3rd Qu.:0.0022176 3rd Qu.:0.0006001
## Max. :0.0006537 Max. :0.051902 Max. :0.0038765 Max. :0.0009875
## M22 M23 M24
## Min. :0.0001147 Min. :4.095e-06 Min. :0.0002164
## 1st Qu.:0.0001568 1st Qu.:2.040e-05 1st Qu.:0.0003963
## Median :0.0002418 Median :3.146e-05 Median :0.0004330
## Mean :0.0002434 Mean :4.158e-05 Mean :0.0004347
## 3rd Qu.:0.0002925 3rd Qu.:5.554e-05 3rd Qu.:0.0004681
## Max. :0.0004692 Max. :1.300e-04 Max. :0.0006875
## M25 M26 M27 M28
## Min. :0.0001040 Min. :0.001629 Min. :0.02285 Min. :0.0001008
## 1st Qu.:0.0002203 1st Qu.:0.002172 1st Qu.:0.04220 1st Qu.:0.0002034
## Median :0.0002558 Median :0.002947 Median :0.05243 Median :0.0002291
## Mean :0.0002574 Mean :0.003003 Mean :0.05299 Mean :0.0002460
## 3rd Qu.:0.0002883 3rd Qu.:0.003554 3rd Qu.:0.05962 3rd Qu.:0.0002698
## Max. :0.0003942 Max. :0.006034 Max. :0.08730 Max. :0.0004238
## M29 M30 groupe
## Min. :3.515e-05 Min. :0.0001366 A1:12
## 1st Qu.:5.891e-05 1st Qu.:0.0002730 B1:12
## Median :7.035e-05 Median :0.0003661 A2: 8
## Mean :7.322e-05 Mean :0.0003574 B2: 6
## 3rd Qu.:8.831e-05 3rd Qu.:0.0004212
## Max. :1.158e-04 Max. :0.0005836head(metabolite)## M1 M2 M3 M4 M5 M6
## 1 0.0019497303 0.0009976818 0.005072411 0.001583864 0.002111279 0.001022931
## 2 0.0021994583 0.0004305712 0.007059411 0.001773148 0.002922657 0.001118970
## 3 0.0019755328 0.0015505552 0.006012389 0.002276433 0.003616820 0.001564363
## 4 0.0006109359 0.0004927399 0.010356908 0.002214662 0.001517629 0.001285894
## 5 0.0015460363 0.0019539049 0.011131809 0.002525373 0.001881233 0.001228354
## 6 0.0018525065 0.0011651985 0.006663383 0.002733360 0.005809555 0.001249846
## M7 M8 M9 M10 M11 M12
## 1 0.0001088416 6.670640e-05 0.001797589 0.002438347 0.0003045516 0.003647887
## 2 0.0001090255 7.117895e-05 0.004543185 0.002179489 0.0004631070 0.004411536
## 3 0.0003023043 1.056315e-04 0.002996306 0.002126223 0.0003786612 0.005402707
## 4 0.0001507176 2.070830e-04 0.016050478 0.002033345 0.0006182604 0.004830741
## 5 0.0002611004 1.551742e-04 0.022257759 0.002332064 0.0006151416 0.006485944
## 6 0.0002125418 1.479881e-04 0.003708689 0.002743453 0.0004567703 0.005051771
## M13 M14 M15 M16 M17 M18
## 1 0.0003550040 0.0002140638 0.0003044825 0.0003427589 0.0003572941 0.0002825989
## 2 0.0005174194 0.0001826632 0.0002804458 0.0002286762 0.0004739388 0.0002522240
## 3 0.0004331546 0.0002477032 0.0003259931 0.0001055723 0.0004826770 0.0003256138
## 4 0.0003652183 0.0002564543 0.0007757909 0.0004552280 0.0005090956 0.0002841940
## 5 0.0006020554 0.0002922910 0.0005687291 0.0011965096 0.0007846011 0.0005147707
## 6 0.0004601564 0.0003145901 0.0003627486 0.0005267340 0.0007029728 0.0006537269
## M19 M20 M21 M22 M23 M24
## 1 0.01069387 0.001262816 0.0003511469 0.0001487752 2.530329e-05 0.0003748787
## 2 0.01270249 0.001370909 0.0004284019 0.0002725926 2.384110e-05 0.0004172943
## 3 0.01177159 0.001452808 0.0005482662 0.0002346421 1.775549e-05 0.0004433637
## 4 0.03503002 0.001225808 0.0005115621 0.0002712070 3.131458e-05 0.0004172762
## 5 0.05190167 0.002617515 0.0005729502 0.0003475241 4.095478e-06 0.0006517901
## 6 0.01661872 0.002282455 0.0007620550 0.0003296081 2.477516e-05 0.0005478172
## M25 M26 M27 M28 M29 M30
## 1 0.0001721437 0.002133350 0.04148790 0.0001733488 4.308620e-05 0.0002326818
## 2 0.0001924611 0.002460116 0.05349343 0.0001876042 7.318912e-05 0.0002124977
## 3 0.0002629841 0.002034839 0.05460285 0.0001825373 6.950998e-05 0.0002901140
## 4 0.0002217138 0.003712321 0.04205769 0.0002708714 8.408064e-05 0.0002720904
## 5 0.0003787150 0.004648184 0.06862181 0.0003919751 1.141668e-04 0.0005107015
## 6 0.0003942078 0.002749467 0.05497483 0.0004238065 7.366480e-05 0.0005836174
## groupe
## 1 A1
## 2 A1
## 3 A1
## 4 A1
## 5 A1
## 6 A1as_tibble(metabolite)## # A tibble: 38 × 31
## M1 M2 M3 M4 M5 M6 M7 M8 M9
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 0.00195 0.000998 0.00507 0.00158 0.00211 0.00102 0.000109 0.0000667 0.00180
## 2 0.00220 0.000431 0.00706 0.00177 0.00292 0.00112 0.000109 0.0000712 0.00454
## 3 0.00198 0.00155 0.00601 0.00228 0.00362 0.00156 0.000302 0.000106 0.00300
## 4 0.000611 0.000493 0.0104 0.00221 0.00152 0.00129 0.000151 0.000207 0.0161
## 5 0.00155 0.00195 0.0111 0.00253 0.00188 0.00123 0.000261 0.000155 0.0223
## 6 0.00185 0.00117 0.00666 0.00273 0.00581 0.00125 0.000213 0.000148 0.00371
## 7 0.000525 0.000881 0.00849 0.00156 0.00155 0.000760 0.000109 0.000116 0.00938
## 8 0.00769 0.00292 0.00831 0.00322 0.00402 0.00176 0.000309 0.000277 0.00905
## 9 0.00361 0.00314 0.00970 0.00263 0.00507 0.00208 0.000207 0.000297 0.00816
## 10 0.00305 0.00232 0.00760 0.00254 0.00490 0.00270 0.000256 0.000215 0.00466
## # … with 28 more rows, and 22 more variables: M10 <dbl>, M11 <dbl>, M12 <dbl>,
## # M13 <dbl>, M14 <dbl>, M15 <dbl>, M16 <dbl>, M17 <dbl>, M18 <dbl>,
## # M19 <dbl>, M20 <dbl>, M21 <dbl>, M22 <dbl>, M23 <dbl>, M24 <dbl>,
## # M25 <dbl>, M26 <dbl>, M27 <dbl>, M28 <dbl>, M29 <dbl>, M30 <dbl>,
## # groupe <fct>table(metabolite$groupe)##
## A1 B1 A2 B2
## 12 12 8 6Représentation souhaitée (pour un seul métabolite)
Représentation de boxplots (avec les points) par groupe avec p-value (ou symbole) d’un test de comparaison entre groupes.
Utilisons le package ggpubr dont le “slogan” est
‘ggplot2’ Based Publication Ready Plots.
library(ggpubr)Juste les boxplots
Boxplot_ggpubr <- ggboxplot(data = metabolite,
x = "groupe", y = "M1",
color = "groupe")
Boxplot_ggpubr
Avec les points
Boxplot_ggpubr <- ggboxplot(data = metabolite,
x = "groupe", y = "M1",
color = "groupe",
add="jitter")
Boxplot_ggpubr
Jusqu’ici on pouvait faire la même chose directement avec
ggplot2 à un thème prêt.
ggplot(metabolite, aes(x = groupe, y = M1, color = groupe)) + geom_boxplot() + geom_jitter(width=0.1) +
theme(legend.position = "top")
Mais ce qui est souhaité est (presque) ceci avec un test de Kruskal-Wallis de comparaison des 4 groupes :
Boxplot_ggpubr + stat_compare_means()
Pour une ANOVA :
Boxplot_ggpubr + stat_compare_means(method = "anova")
et finalement ceci avec les comparaisons deux-à-deux effectuée par un test de Wilcoxon :
MesComparaisons <- list( c("A1", "B1"),
c("A2", "B2"),
c("A1", "A2"),
c("B1", "B2"))
Boxplot_ggpubr + stat_compare_means(comparisons = MesComparaisons)
Pour réaliser des tests de Student :
Boxplot_ggpubr + stat_compare_means(comparisons = MesComparaisons,
method = "t.test")
Pour afficher à la fois test global et comparaison deux-à-deux (avec un petit défaut dans l’affichage) :
Boxplot_ggpubr +
stat_compare_means(comparisons = MesComparaisons) +
stat_compare_means()
Représenter tous les métabolites
Nécessite de modifier la structure du jeu de données qui était pourtant tidy.
La tidyness ultime reviendrait-elle à rassembler toutes les données numériques en une seule colonne ?
pivot_longer
metabolite_longer <- tidyr::pivot_longer(metabolite, !groupe,
names_to="Metab", values_to = "Abundance")
dim(metabolite_longer)## [1] 1140 3metabolite_longer## # A tibble: 1,140 × 3
## groupe Metab Abundance
## <fct> <chr> <dbl>
## 1 A1 M1 0.00195
## 2 A1 M2 0.000998
## 3 A1 M3 0.00507
## 4 A1 M4 0.00158
## 5 A1 M5 0.00211
## 6 A1 M6 0.00102
## 7 A1 M7 0.000109
## 8 A1 M8 0.0000667
## 9 A1 M9 0.00180
## 10 A1 M10 0.00244
## # … with 1,130 more rowssummary(metabolite_longer)## groupe Metab Abundance
## A1:360 Length:1140 Min. :0.0000041
## B1:360 Class :character 1st Qu.:0.0002711
## A2:240 Mode :character Median :0.0005297
## B2:180 Mean :0.0037821
## 3rd Qu.:0.0023096
## Max. :0.0872987table(metabolite_longer$Metab)##
## M1 M10 M11 M12 M13 M14 M15 M16 M17 M18 M19 M2 M20 M21 M22 M23 M24 M25 M26 M27
## 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38
## M28 M29 M3 M30 M4 M5 M6 M7 M8 M9
## 38 38 38 38 38 38 38 38 38 38Graphique de base
Avec ggpubr
ggboxplot(data = metabolite_longer,
x = "groupe", y = "Abundance", color = "groupe",
facet.by = "Metab")
Pour libérer les échelles verticales et ajouter les points en plus des boxplots :
Boxplot_ggpubr_tous <- ggboxplot(data = metabolite_longer,
x = "groupe", y = "Abundance",
color = "groupe",
facet.by = "Metab",
scales = "free_y",
add = "jitter")
Boxplot_ggpubr_tous
Pour ajouter un symbole de significativité :
Boxplot_ggpubr_tous + stat_compare_means(comparisons = MesComparaisons,
label = "p.signif")
Avec ggplot2
Boxplot + points avec groupe en abscisse et en couleur
et Abundance en ordonnée.
Boxplot_gg2 <- ggplot(data = metabolite_longer,
aes(x = groupe, y = Abundance, color = groupe)) +
geom_boxplot() + geom_jitter(width=0.1)facet_wrap
Boxplot_gg2 + facet_wrap(~Metab)
Comme précedemment, on va là aussi libérer les échelles verticales.
facet_wrap avec scales = "free_y"
Pour relacher cette contrainte :
Boxplot_gg2 + facet_wrap(~Metab, scales = "free_y")
On va voir maintenant comment ajouter les p-values ou les symboles de significativité des tests et des segments ou autres pour indiquer les paires testées.
Importation des p-values
Les p-values sont dans un fichier à part.
Pvalues <- read.xlsx("Pvalues.xlsx", sheet=1)
Pvalues## group1 group2 M1 M2 M3 M4 M5 M6 M7
## 1 B1 A1 0.93099 0.370840 0.099877 0.12602 0.26024 1.000000 0.37084
## 2 B2 A2 0.94972 0.490840 0.572760 0.66200 0.22844 0.662000 0.66200
## 3 A1 A2 0.20302 0.058729 0.562830 0.20302 0.61603 0.082579 0.37495
## 4 B1 B2 1.00000 0.437080 0.891620 0.89162 0.75027 0.437080 0.43708
## M8 M9 M10 M11 M12 M13 M14 M15 M16
## 1 0.58336 0.37084 0.97697 0.97697 0.402500 0.140960 0.31232 0.2854800 0.54437
## 2 0.75458 0.14186 0.49084 0.85181 0.754580 0.754580 0.41359 0.5727600 0.49084
## 3 0.84705 0.37495 0.46359 0.56283 0.049141 0.027891 0.15349 0.0038135 0.90787
## 4 0.68197 0.29078 0.29078 0.43708 0.249620 0.891620 0.29078 0.1505100 0.10246
## M17 M18 M19 M20 M21 M22 M23 M24 M25
## 1 0.31232 0.93099 0.43573 0.79501 0.83986 0.0463870 0.70745 0.28548 0.58336
## 2 0.57276 0.57276 0.34499 0.18115 0.22844 0.9497200 0.75458 0.85181 0.41359
## 3 0.33485 0.78714 0.29762 0.72845 0.90787 0.0038135 0.20302 0.84705 0.61603
## 4 0.43708 0.43708 0.38451 0.15051 0.38451 0.2907800 0.38451 0.82008 0.43708
## M26 M27 M28 M29 M30
## 1 0.28548 0.236580 0.70745 0.19393 0.83986
## 2 0.85181 0.490840 0.85181 0.66200 0.34499
## 3 0.15349 0.097164 0.67132 0.17697 1.00000
## 4 0.96359 0.249620 0.49364 0.68197 0.38451On va préparer un nouveau data.frame df_annotation sur
la base de ces p-values.
Créer les symbôles de significativité
symbol <- apply(Pvalues[,-c(1,2)], 1,
function(x){
symnum(x,
cutpoints = c(0,.001,.01,.05, .1, 1),
symbols = c("****","***","**","*","."))})
df_annotation <- as.data.frame(symbol)
colnames(df_annotation) <- c("B1_A1","B2_A2","A1_A2","B1_B2")
df_annotation <- rownames_to_column(df_annotation,"Metab")
df_annotation## Metab B1_A1 B2_A2 A1_A2 B1_B2
## 1 M1 . . . .
## 2 M2 . . * .
## 3 M3 * . . .
## 4 M4 . . . .
## 5 M5 . . . .
## 6 M6 . . * .
## 7 M7 . . . .
## 8 M8 . . . .
## 9 M9 . . . .
## 10 M10 . . . .
## 11 M11 . . . .
## 12 M12 . . ** .
## 13 M13 . . ** .
## 14 M14 . . . .
## 15 M15 . . *** .
## 16 M16 . . . .
## 17 M17 . . . .
## 18 M18 . . . .
## 19 M19 . . . .
## 20 M20 . . . .
## 21 M21 . . . .
## 22 M22 ** . *** .
## 23 M23 . . . .
## 24 M24 . . . .
## 25 M25 . . . .
## 26 M26 . . . .
## 27 M27 . . * .
## 28 M28 . . . .
## 29 M29 . . . .
## 30 M30 . . . .Où placer les p-values ou les symbôles ?
Attention bricolage à venir !
Pour placer les p-values au dessus des boxplots (et des points), il faut faire de la place ! Commençons par repérer le mininum et le maximum de chaque métabolite
df_annotation$min <- apply(metabolite[,1:30], 2, min)
df_annotation$max <- apply(metabolite[,1:30], 2, max)
df_annotation## Metab B1_A1 B2_A2 A1_A2 B1_B2 min max
## 1 M1 . . . . 5.253883e-04 0.0076898600
## 2 M2 . . * . 3.556876e-04 0.0031358399
## 3 M3 * . . . 2.925137e-03 0.0130006779
## 4 M4 . . . . 1.128671e-03 0.0035038216
## 5 M5 . . . . 9.152375e-04 0.0058095546
## 6 M6 . . * . 6.342488e-04 0.0042976851
## 7 M7 . . . . 6.880382e-05 0.0003651556
## 8 M8 . . . . 6.201652e-05 0.0003739061
## 9 M9 . . . . 6.150863e-04 0.0222577590
## 10 M10 . . . . 9.801484e-04 0.0042863095
## 11 M11 . . . . 2.613533e-04 0.0012366796
## 12 M12 . . ** . 2.135441e-03 0.0064859444
## 13 M13 . . ** . 2.097845e-04 0.0010539301
## 14 M14 . . . . 1.195333e-04 0.0005065246
## 15 M15 . . *** . 1.814674e-04 0.0007757909
## 16 M16 . . . . 2.290687e-05 0.0011965096
## 17 M17 . . . . 2.101212e-04 0.0009212711
## 18 M18 . . . . 1.488194e-04 0.0006537269
## 19 M19 . . . . 6.853101e-03 0.0519016723
## 20 M20 . . . . 7.241679e-04 0.0038764847
## 21 M21 . . . . 1.916745e-04 0.0009875093
## 22 M22 ** . *** . 1.147069e-04 0.0004692441
## 23 M23 . . . . 4.095478e-06 0.0001300232
## 24 M24 . . . . 2.163686e-04 0.0006875205
## 25 M25 . . . . 1.040086e-04 0.0003942078
## 26 M26 . . . . 1.628724e-03 0.0060340080
## 27 M27 . . * . 2.284627e-02 0.0872986452
## 28 M28 . . . . 1.008304e-04 0.0004238065
## 29 M29 . . . . 3.514620e-05 0.0001158070
## 30 M30 . . . . 1.365880e-04 0.0005836174Et calculons les coordonnées des positions des p-values en prenant en ordonnée une valeur légèrement supérieure au maximum de chaque métabolite.
df_annotation$x_pval_1 <- 1.5
df_annotation$y_pval_1 <- 1.15*df_annotation$max
df_annotation$x_pval_2 <- 3.5
df_annotation$y_pval_2 <- 1.15*df_annotation$max
df_annotation$x_pval_3 <- 2
df_annotation$y_pval_3 <- 1.25*df_annotation$max
df_annotation$x_pval_4 <- 3
df_annotation$y_pval_4 <- 1.35*df_annotation$maxBoxplot_gg2_Annotate <- ggplot(data = metabolite_longer,
aes(x=groupe, y=Abundance, color=groupe)) +
geom_boxplot() +
facet_wrap(~Metab, scales = "free_y") +
geom_jitter(width=0.1) +
geom_text(data = df_annotation,
aes(x = x_pval_1, y = y_pval_1,
label = B1_A1), color="black", size=5) +
geom_text(data = df_annotation,
aes(x = x_pval_2, y = y_pval_2,
label = B2_A2), color="black", size=5) +
geom_text(data = df_annotation,
aes(x = x_pval_3, y = y_pval_3,
label = A1_A2), color="black", size=5) +
geom_text(data = df_annotation,
aes(x = x_pval_4, y = y_pval_4,
label = B1_B2), color="black", size=5)
Boxplot_gg2_Annotate
Ajouter les segments
- Compléter
df_annotationavec les coordonnées des extrémités des segments
df_annotation$x_bar_1 <- 1
df_annotation$y_bar_1 <- 1.13*df_annotation$max
df_annotation$x_end_bar_1 <- 2
df_annotation$y_end_bar_1 <- 1.13*df_annotation$max
df_annotation$x_bar_2 <- 3
df_annotation$y_bar_2 <- 1.13*df_annotation$max
df_annotation$x_end_bar_2 <- 4
df_annotation$y_end_bar_2 <- 1.13*df_annotation$max
df_annotation$x_bar_3 <- 1
df_annotation$y_bar_3 <- 1.23*df_annotation$max
df_annotation$x_end_bar_3 <- 3
df_annotation$y_end_bar_3 <- 1.23*df_annotation$max
df_annotation$x_bar_4 <- 2
df_annotation$y_bar_4 <- 1.33*df_annotation$max
df_annotation$x_end_bar_4 <- 4
df_annotation$y_end_bar_4 <- 1.33*df_annotation$max- Ajouter les segments avec
geom_segment
Boxplot_gg2_Annotate +
geom_segment(data = df_annotation,
aes(x = x_bar_1, y = y_bar_1,
xend = x_end_bar_1, yend = y_end_bar_1), color="black") +
geom_segment(data = df_annotation,
aes(x = x_bar_2, y = y_bar_2,
xend = x_end_bar_2, yend = y_end_bar_2), color="black") +
geom_segment(data = df_annotation,
aes(x = x_bar_3, y = y_bar_3,
xend = x_end_bar_3, yend = y_end_bar_3), color="black") +
geom_segment(data = df_annotation,
aes(x = x_bar_4, y = y_bar_4,
xend = x_end_bar_4, yend = y_end_bar_4), color="black")
- Cela pourrait aussi être des courbes avec
geom_curve
Boxplot_gg2_Annotate +
geom_curve(data = df_annotation,
aes(x = x_bar_1, y = y_bar_1,
xend = x_end_bar_1, yend = y_end_bar_1),
color="black", curvature = -0.5) +
geom_curve(data = df_annotation,
aes(x = x_bar_2, y = y_bar_2,
xend = x_end_bar_2, yend = y_end_bar_2),
color="black", curvature = -0.5) +
geom_curve(data = df_annotation,
aes(x = x_bar_3, y = y_bar_3,
xend = x_end_bar_3, yend = y_end_bar_3),
color="black", curvature = -0.5) +
geom_curve(data = df_annotation,
aes(x = x_bar_4, y = y_bar_4,
xend = x_end_bar_4, yend = y_end_bar_4),
color="black", curvature = -0.5)
- Ajuster l’échelle verticale
Pour que la courbe ne sorte pas du graphique, on peut réaliser un
dernier petit ajustement avec geom_blank et un nouveau
data.frame (et/ou modifier la courbure).
df_dummy <- select(df_annotation, Metab, min, max)
df_dummy$groupe <- "A1"
df_dummy$min <- 0.9*df_dummy$min
df_dummy$max <- 1.5*df_dummy$maxBoxplot_gg2_Annotate +
geom_curve(data = df_annotation,
aes(x = x_bar_1, y = y_bar_1,
xend = x_end_bar_1, yend = y_end_bar_1),
color="black", curvature = -0.25) +
geom_curve(data = df_annotation,
aes(x = x_bar_2, y = y_bar_2,
xend = x_end_bar_2, yend = y_end_bar_2),
color="black", curvature = -0.25) +
geom_curve(data = df_annotation,
aes(x = x_bar_3, y = y_bar_3,
xend = x_end_bar_3, yend = y_end_bar_3),
color="black", curvature = -0.25) +
geom_curve(data = df_annotation,
aes(x = x_bar_4, y = y_bar_4,
xend = x_end_bar_4, yend = y_end_bar_4),
color="black", curvature = -0.25) +
geom_blank(data = df_dummy, aes(x = groupe, y = max)) +
geom_blank(data = df_dummy, aes(x = groupe, y = min))
En résumé
ggpubrpourquoi pas.quelques découvertes
ggplot2:scales = "free_y",geom_curve,geom_blank, manipulation de plusieurs data.frames pour un seul graphique.plus
tidyquetidy?
Bonus : barplot + barres d’erreur
df_bar <- summarise(group_by(metabolite_longer, groupe, Metab),
mean=mean(Abundance),
stdev=sd(Abundance))
Barplot_ggplot2 <- ggplot(df_bar) +
geom_bar(aes(x=groupe, y=mean, fill = groupe), stat="identity", alpha=0.5) +
geom_errorbar(aes(x=groupe, ymin=mean-stdev, ymax=mean+stdev),
size=.3, width=.3) +
facet_wrap(~Metab, scales = "free_y") +
geom_jitter(data=metabolite_longer, aes(x=groupe, y=Abundance, color=groupe),
width=0.1, size=1, alpha=0.3) +
scale_colour_manual(values = c("rosybrown","tomato","midnightblue", "purple")) +
scale_fill_manual(values = c("rosybrown","tomato","midnightblue", "purple")) +
theme(strip.text = element_text(face="bold")) +
geom_text(data = df_annotation, aes(x = x_pval_1, y = y_pval_1,
label = B1_A1),
color="black", size=3) +
geom_text(data = df_annotation, aes(x = x_pval_2, y = y_pval_2,
label = B2_A2),
color="black", size=3) +
geom_text(data = df_annotation, aes(x = x_pval_3, y = y_pval_3,
label = A1_A2),
color="black", size=3) +
geom_text(data = df_annotation, aes(x = x_pval_4, y = y_pval_4,
label = B1_B2),
color="black", size=3) +
geom_segment(data = df_annotation,
aes(x = x_bar_1, y = y_bar_1,
xend = x_end_bar_1, yend = y_end_bar_1),
color="black") +
geom_segment(data = df_annotation,
aes(x = x_bar_2, y = y_bar_2,
xend = x_end_bar_2, yend = y_end_bar_2),
color="black") +
geom_segment(data = df_annotation,
aes(x = x_bar_3, y = y_bar_3,
xend = x_end_bar_3, yend = y_end_bar_3),
color="black") +
geom_segment(data = df_annotation,
aes(x = x_bar_4, y = y_bar_4,
xend = x_end_bar_4, yend = y_end_bar_4),
color="black")
Barplot_ggplot2