setwd("C:/Users/laura/OneDrive/Escritorio/Doctorado/Ari/Análisis")
#Librerías
library(ggplot2)
library(ggthemes)
library(car)
library(effsize)
library(knitr)
library(rmarkdown)
# Cargar datos
Datos<-read.csv("AnalisisR.csv", header=TRUE)
# 2. Convertir ambas columnas a Factores
Datos$Modelo <- factor(Datos$Modelo)
Datos$Tratamiento <- factor(Datos$Tratamiento)
################################################################################
#Junctions
ggplot(data = Datos, aes(x = factor(Modelo), y = Junctions, fill=factor(Modelo))) +
scale_fill_manual(values=c("#9E9E9E", "#009688")) +
geom_boxplot(outlier.color="red") +
facet_wrap(~Tratamiento) +
geom_jitter(width = 0.1, size = 1, color = "black") +
labs(title="Branch junctions density ",
x="", y="junctions/μm2") +
theme_few()+
labs(fill="Modelo")
#Prueba de normalidad
hist(Datos$Junctions, main = "Branch junctions density",
ylab = "Frecuencia", xlab = "junctions/μm2", col = "lightpink")
shapiro.test(lm(Junctions~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(Junctions ~ Muestra, Datos)$residuals
## W = 0.94969, p-value = 0.4847
by(Datos$Junctions, Datos$Muestra, shapiro.test)
## Datos$Muestra: Glut
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.9555, p-value = 0.7506
##
## ----------------------------------------------------------------
## Datos$Muestra: Glut-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.74993, p-value = 0.03857
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.89966, p-value = 0.4294
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.98307, p-value = 0.9198
#Los datos se distribuyen normalmente
#Homoscedasticidad
bartlett.test(Junctions~Muestra, Datos)
##
## Bartlett test of homogeneity of variances
##
## data: Junctions by Muestra
## Bartlett's K-squared = 5.3455, df = 3, p-value = 0.1482
#Hay homoscedasticidad de varianza
# ANOVA de 2 Vías con interacción
modelo_anova <- aov(Junctions ~ Modelo * Tratamiento, data = Datos)
summary(modelo_anova)
## Df Sum Sq Mean Sq F value Pr(>F)
## Modelo 1 3.040e-06 3.035e-06 0.771 0.397
## Tratamiento 1 6.660e-06 6.661e-06 1.692 0.218
## Modelo:Tratamiento 1 1.167e-05 1.167e-05 2.965 0.111
## Residuals 12 4.724e-05 3.937e-06
# Post-hoc de Tukey
TukeyHSD(modelo_anova)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = Junctions ~ Modelo * Tratamiento, data = Datos)
##
## $Modelo
## diff lwr upr p adj
## sham-glut 0.0008710813 -0.001290466 0.003032628 0.3971623
##
## $Tratamiento
## diff lwr upr p adj
## vehiculo-urb 0.001290481 -0.0008710658 0.003452028 0.2177554
##
## $`Modelo:Tratamiento`
## diff lwr upr p adj
## sham:urb-glut:urb 0.0025792350 -0.001586153 0.006744623 0.3034764
## glut:vehiculo-glut:urb 0.0029986350 -0.001166753 0.007164023 0.1964371
## sham:vehiculo-glut:urb 0.0021615625 -0.002003826 0.006326951 0.4453733
## glut:vehiculo-sham:urb 0.0004194000 -0.003745988 0.004584788 0.9902347
## sham:vehiculo-sham:urb -0.0004176725 -0.004583061 0.003747716 0.9903516
## sham:vehiculo-glut:vehiculo -0.0008370725 -0.005002461 0.003328316 0.9311541
########################################################################
#Length
ggplot(data = Datos, aes(x = factor(Modelo), y = Length, fill=factor(Modelo))) +
scale_fill_manual(values=c("#9E9E9E", "#009688")) +
geom_boxplot(outlier.color="red") +
facet_wrap(~Tratamiento) +
geom_jitter(width = 0.1, size = 1, color = "black") +
labs(title="Average branch length",
x="", y="Average branch length (μm)") +
theme_few()+
labs(fill="Modelo")
#Prueba de normalidad
hist(Datos$Length, main = "Average branch length",
ylab = "Frecuencia", xlab = "Average branch length (μm)", col = "lightpink")
shapiro.test(lm(Length~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(Length ~ Muestra, Datos)$residuals
## W = 0.97335, p-value = 0.8895
by(Datos$Length, Datos$Muestra, shapiro.test)
## Datos$Muestra: Glut
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.85737, p-value = 0.2509
##
## ----------------------------------------------------------------
## Datos$Muestra: Glut-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.91042, p-value = 0.4847
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.87498, p-value = 0.3176
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.89057, p-value = 0.3858
#Los datos se distribuyen normalmente
#Homoscedasticidad
bartlett.test(Length~Muestra, Datos)
##
## Bartlett test of homogeneity of variances
##
## data: Length by Muestra
## Bartlett's K-squared = 3.3253, df = 3, p-value = 0.3441
#Hay homoscedasticidad de varianza
# ANOVA de 2 Vías con interacción
modelo_anova2 <- aov(Length ~ Modelo * Tratamiento, data = Datos)
summary(modelo_anova2)
## Df Sum Sq Mean Sq F value Pr(>F)
## Modelo 1 8.152 8.152 5.813 0.0329 *
## Tratamiento 1 4.853 4.853 3.461 0.0875 .
## Modelo:Tratamiento 1 4.053 4.053 2.890 0.1149
## Residuals 12 16.828 1.402
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
# Post-hoc de Tukey
TukeyHSD(modelo_anova2)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = Length ~ Modelo * Tratamiento, data = Datos)
##
## $Modelo
## diff lwr upr p adj
## sham-glut 1.427573 0.1374789 2.717667 0.0328529
##
## $Tratamiento
## diff lwr upr p adj
## vehiculo-urb -1.101476 -2.391569 0.1886185 0.0875236
##
## $`Modelo:Tratamiento`
## diff lwr upr p adj
## sham:urb-glut:urb 2.43411618 -0.05194694 4.92017929 0.0556397
## glut:vehiculo-glut:urb -0.09493216 -2.58099527 2.39113096 0.9994480
## sham:vehiculo-glut:urb 0.32609733 -2.15996579 2.81216044 0.9790175
## glut:vehiculo-sham:urb -2.52904833 -5.01511145 -0.04298522 0.0457533
## sham:vehiculo-sham:urb -2.10801885 -4.59408196 0.37804426 0.1073264
## sham:vehiculo-glut:vehiculo 0.42102948 -2.06503363 2.90709260 0.9568605
##############################################################################
#MaxBL
ggplot(data = Datos, aes(x = factor(Modelo), y = MaxBL, fill=factor(Modelo))) +
scale_fill_manual(values=c("#9E9E9E", "#009688")) +
geom_boxplot(outlier.color="red") +
facet_wrap(~Tratamiento) +
geom_jitter(width = 0.1, size = 1, color = "black") +
labs(title="Maximum branch length",
x="", y="Maximum branch length (μm)") +
theme_few()+
labs(fill="Modelo")
#Prueba de normalidad
hist(Datos$MaxBL, main = "Maximum branch length",
ylab = "Frecuencia", xlab = "Maximum branch length (μm)", col = "lightpink")
shapiro.test(lm(MaxBL~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(MaxBL ~ Muestra, Datos)$residuals
## W = 0.95876, p-value = 0.6394
by(Datos$MaxBL, Datos$Muestra, shapiro.test)
## Datos$Muestra: Glut
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.90629, p-value = 0.463
##
## ----------------------------------------------------------------
## Datos$Muestra: Glut-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.90938, p-value = 0.4791
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.93292, p-value = 0.6117
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.97471, p-value = 0.8704
#Los datos se distribuyen normalmente
#Homoscedasticidad
bartlett.test(MaxBL~Muestra, Datos)
##
## Bartlett test of homogeneity of variances
##
## data: MaxBL by Muestra
## Bartlett's K-squared = 1.894, df = 3, p-value = 0.5947
#Hay homoscedasticidad de varianza
# ANOVA de 2 Vías con interacción
modelo_anova3 <- aov(MaxBL ~ Modelo * Tratamiento, data = Datos)
summary(modelo_anova3)
## Df Sum Sq Mean Sq F value Pr(>F)
## Modelo 1 14.76 14.761 6.387 0.0265 *
## Tratamiento 1 10.37 10.372 4.488 0.0557 .
## Modelo:Tratamiento 1 17.10 17.098 7.399 0.0186 *
## Residuals 12 27.73 2.311
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
# Post-hoc de Tukey
TukeyHSD(modelo_anova3)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = MaxBL ~ Modelo * Tratamiento, data = Datos)
##
## $Modelo
## diff lwr upr p adj
## sham-glut 1.920983 0.2648946 3.577072 0.0265499
##
## $Tratamiento
## diff lwr upr p adj
## vehiculo-urb -1.610272 -3.266361 0.04581685 0.05568
##
## $`Modelo:Tratamiento`
## diff lwr upr p adj
## sham:urb-glut:urb 3.9884860 0.7971359 7.1798361 0.0136892
## glut:vehiculo-glut:urb 0.4572305 -2.7341196 3.6485807 0.9730198
## sham:vehiculo-glut:urb 0.3107114 -2.8806387 3.5020615 0.9911472
## glut:vehiculo-sham:urb -3.5312555 -6.7226056 -0.3399054 0.0288442
## sham:vehiculo-sham:urb -3.6777746 -6.8691247 -0.4864245 0.0227187
## sham:vehiculo-glut:vehiculo -0.1465191 -3.3378692 3.0448310 0.9990432
################################################################################
#TotalPL
ggplot(data = Datos, aes(x = factor(Modelo), y = TotalPL, fill=factor(Modelo))) +
scale_fill_manual(values=c("#9E9E9E", "#009688")) +
geom_boxplot(outlier.color="red") +
facet_wrap(~Tratamiento) +
geom_jitter(width = 0.1, size = 1, color = "black") +
labs(title="Microglial process density",
x="", y="Microglial process density (μm/μm²)") +
theme_few()+
labs(fill="Modelo")
#Prueba de normalidad
hist(Datos$TotalPL, main = "Microglial process density",
ylab = "Frecuencia", xlab = "Microglial process density (μm/μm²)", col = "lightpink")
shapiro.test(lm(TotalPL~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(TotalPL ~ Muestra, Datos)$residuals
## W = 0.9725, p-value = 0.8773
by(Datos$TotalPL, Datos$Muestra, shapiro.test)
## Datos$Muestra: Glut
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.92915, p-value = 0.5894
##
## ----------------------------------------------------------------
## Datos$Muestra: Glut-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.76237, p-value = 0.05008
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.8924, p-value = 0.3943
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.9175, p-value = 0.523
#Los datos se distribuyen normalmente
#Homoscedasticidad
bartlett.test(TotalPL~Muestra, Datos)
##
## Bartlett test of homogeneity of variances
##
## data: TotalPL by Muestra
## Bartlett's K-squared = 5.1821, df = 3, p-value = 0.1589
#Hay homoscedasticidad de varianza
# ANOVA de 2 Vías con interacción
modelo_anova4 <- aov(TotalPL ~ Modelo * Tratamiento, data = Datos)
summary(modelo_anova4)
## Df Sum Sq Mean Sq F value Pr(>F)
## Modelo 1 0.000638 0.000638 0.769 0.3976
## Tratamiento 1 0.000150 0.000150 0.180 0.6785
## Modelo:Tratamiento 1 0.007135 0.007135 8.602 0.0125 *
## Residuals 12 0.009953 0.000829
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
# Post-hoc de Tukey
TukeyHSD(modelo_anova4)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = TotalPL ~ Modelo * Tratamiento, data = Datos)
##
## $Modelo
## diff lwr upr p adj
## sham-glut 0.01263102 -0.01874295 0.04400499 0.3976103
##
## $Tratamiento
## diff lwr upr p adj
## vehiculo-urb 0.006117424 -0.02525655 0.03749139 0.6784759
##
## $`Modelo:Tratamiento`
## diff lwr upr p adj
## sham:urb-glut:urb 0.054864505 -0.005594403 0.11532341 0.0799335
## glut:vehiculo-glut:urb 0.048350910 -0.012107998 0.10880982 0.1354814
## sham:vehiculo-glut:urb 0.018748443 -0.041710465 0.07920735 0.7945315
## glut:vehiculo-sham:urb -0.006513595 -0.066972503 0.05394531 0.9881079
## sham:vehiculo-sham:urb -0.036116062 -0.096574970 0.02434285 0.3314771
## sham:vehiculo-glut:vehiculo -0.029602467 -0.090061375 0.03085644 0.4925662
###############################################################################
#TotalPLCell
ggplot(data = Datos, aes(x = factor(Modelo), y = TotalPLCell, fill=factor(Modelo))) +
scale_fill_manual(values=c("#9E9E9E", "#009688")) +
geom_boxplot(outlier.color="red") +
facet_wrap(~Tratamiento) +
geom_jitter(width = 0.1, size = 1, color = "black") +
labs(title="Process length per cell",
x="", y="Process length per cell (μm)") +
theme_few()+
labs(fill="Modelo")
#Prueba de normalidad
hist(Datos$TotalPLCell, main = "Process length per cell",
ylab = "Frecuencia", xlab = "Process length per cell (μm)", col = "lightpink")
shapiro.test(lm(TotalPLCell~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(TotalPLCell ~ Muestra, Datos)$residuals
## W = 0.8548, p-value = 0.01605
by(Datos$TotalPLCell, Datos$Muestra, shapiro.test)
## Datos$Muestra: Glut
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.74054, p-value = 0.03137
##
## ----------------------------------------------------------------
## Datos$Muestra: Glut-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.83573, p-value = 0.1832
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.75739, p-value = 0.04519
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.72107, p-value = 0.0199
#Los datos no se distribuyen normalmente
#Homoscedasticidad
bartlett.test(TotalPLCell~Muestra, Datos)
##
## Bartlett test of homogeneity of variances
##
## data: TotalPLCell by Muestra
## Bartlett's K-squared = 4.4589, df = 3, p-value = 0.216
#Hay homoscedasticidad de varianza
Datos$sqrt_TotalPLCell <- sqrt(Datos$TotalPLCell)
shapiro.test(lm(sqrt_TotalPLCell~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(sqrt_TotalPLCell ~ Muestra, Datos)$residuals
## W = 0.8576, p-value = 0.01766
Datos$log_TotalPLCell <- log(Datos$TotalPLCell)
shapiro.test(lm(log_TotalPLCell~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(log_TotalPLCell ~ Muestra, Datos)$residuals
## W = 0.87172, p-value = 0.0289
install.packages("ARTool")
## Warning: package 'ARTool' is in use and will not be installed
library(ARTool)
# 2. Ajustar el modelo no paramétrico de 2 vías
modelo_art <- art(TotalPLCell ~ Modelo * Tratamiento, data = Datos)
# 3. Ver la tabla de ANOVA no paramétrica
anova(modelo_art)
## Analysis of Variance of Aligned Rank Transformed Data
##
## Table Type: Anova Table (Type III tests)
## Model: No Repeated Measures (lm)
## Response: art(TotalPLCell)
##
## Df Df.res F value Pr(>F)
## 1 Modelo 1 12 0.63158 0.4422114
## 2 Tratamiento 1 12 0.60759 0.4507948
## 3 Modelo:Tratamiento 1 12 10.38781 0.0073135 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
# 4. Post-hoc no paramétrico (comparaciones múltiples)
art.con(modelo_art, "Modelo:Tratamiento", adjust="tukey")
## contrast estimate SE df t.ratio p.value
## glut,urb - glut,vehiculo -5.5 2.57 12 -2.144 0.1946
## glut,urb - sham,urb -8.0 2.57 12 -3.118 0.0386
## glut,urb - sham,vehiculo -0.5 2.57 12 -0.195 0.9972
## glut,vehiculo - sham,urb -2.5 2.57 12 -0.974 0.7662
## glut,vehiculo - sham,vehiculo 5.0 2.57 12 1.949 0.2598
## sham,urb - sham,vehiculo 7.5 2.57 12 2.923 0.0541
##
## P value adjustment: tukey method for comparing a family of 4 estimates
##############################################################################
#Microglial
ggplot(data = Datos, aes(x = factor(Modelo), y = Microglial, fill=factor(Modelo))) +
scale_fill_manual(values=c("#9E9E9E", "#009688")) +
geom_boxplot(outlier.color="red") +
facet_wrap(~Tratamiento) +
geom_jitter(width = 0.1, size = 1, color = "black") +
labs(title="Microglial density",
x="", y="Microglial density (cells/cm2)") +
theme_few()+
labs(fill="Modelo")
#Prueba de normalidad
hist(Datos$Microglial, main = "Microglial process density",
ylab = "Frecuencia", xlab = "Microglial process density (μm/μm²)", col = "lightpink")
shapiro.test(lm(Microglial~Muestra, Datos)$residuals)
##
## Shapiro-Wilk normality test
##
## data: lm(Microglial ~ Muestra, Datos)$residuals
## W = 0.97185, p-value = 0.8676
by(Datos$Microglial, Datos$Muestra, shapiro.test)
## Datos$Muestra: Glut
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.95914, p-value = 0.7735
##
## ----------------------------------------------------------------
## Datos$Muestra: Glut-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.96164, p-value = 0.7893
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.8532, p-value = 0.2367
##
## ----------------------------------------------------------------
## Datos$Muestra: Sham-Urb
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.95046, p-value = 0.719
#Los datos se distribuyen normalmente
#Homoscedasticidad
bartlett.test(Microglial~Muestra, Datos)
##
## Bartlett test of homogeneity of variances
##
## data: Microglial by Muestra
## Bartlett's K-squared = 6.2973, df = 3, p-value = 0.09801
#Hay homoscedasticidad de varianza
# ANOVA de 2 Vías con interacción
modelo_anova5 <- aov(Microglial ~ Modelo * Tratamiento, data = Datos)
summary(modelo_anova5)
## Df Sum Sq Mean Sq F value Pr(>F)
## Modelo 1 1.562e+08 1.562e+08 0.270 0.613
## Tratamiento 1 1.447e+09 1.447e+09 2.498 0.140
## Modelo:Tratamiento 1 1.370e+09 1.370e+09 2.364 0.150
## Residuals 12 6.952e+09 5.794e+08
# Post-hoc de Tukey
TukeyHSD(modelo_anova5)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = Microglial ~ Modelo * Tratamiento, data = Datos)
##
## $Modelo
## diff lwr upr p adj
## sham-glut 6249 -19972.88 32470.88 0.6130371
##
## $Tratamiento
## diff lwr upr p adj
## vehiculo-urb 19022.75 -7199.125 45244.63 0.139948
##
## $`Modelo:Tratamiento`
## diff lwr upr p adj
## sham:urb-glut:urb 24752.50 -25778.11 75283.11 0.4922010
## glut:vehiculo-glut:urb 37526.25 -13004.36 88056.86 0.1770711
## sham:vehiculo-glut:urb 25271.75 -25258.86 75802.36 0.4754195
## glut:vehiculo-sham:urb 12773.75 -37756.86 63304.36 0.8747269
## sham:vehiculo-sham:urb 519.25 -50011.36 51049.86 0.9999892
## sham:vehiculo-glut:vehiculo -12254.50 -62785.11 38276.11 0.8872512
#rmarkdown::render("Analisis_Estadisticos.R", output_format = "html_document")
#library(rsconnect)
#setwd("C:/Users/laura/OneDrive/Escritorio/Doctorado/Ari/Análisis")
#result <- rpubsUpload(title='Analisis_Estadisticos',contentFile='Analisis_Estadisticos.html', originalDoc = 'Analisis_Estadisticos.html')
#result