---
title: "Validation_SPARC_Olig2"
author: "Luise A. Seeker"
date: "26/05/2021"
output: html_document
---
```{r}
library(ggplot2)
library(lme4)
library(lmerTest)
library(here)
library(tidyr)
library(webpower)
library(pracma)
```
```{r}
sparc_data <- read.csv(here("data",
"validation_data",
"20220705_variable_classifyer_thresholds_SPARC_RBFOX1_OLIG2.csv"))
```
```{r}
names(sparc_data)
sparc_data$tissue <- as.factor(sparc_data$Tissue)
sparc_data$age_group <- as.factor(sparc_data$age_group)
sparc_data$sex <- as.factor(sparc_data$sex)
sparc_data$total_oligos <- sparc_data$Num_OLIG2 +
sparc_data$Num_RBFOX1_OLIG2+
sparc_data$Num_SPARC_OLIG2+
sparc_data$Num_RBFOX1_SPARC_OLIG2
sparc_data$total_sparc_oligos <-
sparc_data$Num_SPARC_OLIG2+
sparc_data$Num_RBFOX1_SPARC_OLIG2
sparc_data$total_sparc_oligos_perc <- (sparc_data$total_sparc_oligos/
sparc_data$total_oligos) *100
sparc_data$perc_sparc_pos_rbfox1_pos_olig2_pos <- sparc_data$Num_RBFOX1_SPARC_OLIG2/
sparc_data$total_oligos
```
Convert from wide to long format to see how many single, double and triple
positive cells are in each sample
```{r}
keycol <- "sample_id"
valuecol <- "count"
gathercols <- c("Num_OLIG2",
"Num_RBFOX1",
"Num_RBFOX1_OLIG2",
"Num_RBFOX1_SPARC",
"Num_RBFOX1_SPARC_OLIG2",
"Num_SPARC",
"Num_SPARC_OLIG2")
long_data <- gather(sparc_data, keycol, valuecol, gathercols)
head(long_data)
nrow(long_data)
names(long_data)
```
```{r}
long_data$valuecol <- as.numeric(long_data$valuecol)
red_long <- subset(long_data,
long_data$keycol %in% c("Num_OLIG2",
"Num_RBFOX1_OLIG2",
"Num_RBFOX1_SPARC_OLIG2",
"Num_SPARC_OLIG2"))
sort_red_long <- red_long[order(red_long$Tissue),]
ggplot(sort_red_long, aes(x = sample_id ,
y = valuecol,
fill = keycol,
colour = Tissue)) +
geom_bar(position="stack", stat="identity")+ theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
ggplot(sort_red_long, aes(x = sample_id ,
y = valuecol,
fill = keycol)) +
geom_bar(position="stack", stat="identity")+ theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
ggplot(sort_red_long, aes(x = sample_id ,
y = valuecol,
fill = keycol)) +
geom_bar(position="fill", stat="identity")+ theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```
```{r}
ggplot(sparc_data, aes(x = tissue, y = perc_sparc_pos_oligos)) +
geom_boxplot()+ geom_jitter(width = 0.2, aes(colour = sample_id)) +
theme_bw(19) + ylab("SPARC+OLIG2+RBFOX1-HOECHST+ (%)") + xlab("Tissue")
```
```{r}
sparc_data$u_id <- paste(sparc_data$sample_id, sparc_data$tissue, sep = "_")
u_data <- sparc_data[!duplicated(sparc_data$u_id),]
u_data$average_perc_sparc_olig2 <- as.numeric(u_data$average_perc_sparc_olig2)
ggplot(u_data, aes(x = tissue, y = average_perc_sparc_olig2)) +
geom_boxplot()+ geom_jitter(width = 0.2, aes(colour = sample_id)) +
theme_bw(19) + ylab("SPARC+OLIG2+HOECHST+ (%)") + xlab("Tissue")
```
```{r}
ggplot(sparc_data, aes(x = tissue, y = perc_sparc_pos_oligos)) +
geom_boxplot()+ geom_jitter(width = 0.2, aes(colour = sample_id)) +
theme_bw(19) + ylab("SPARC+OLIG2+RBFOX1- as % of OLIG2+") + xlab("Tissue")
```
```{r}
ggplot(sparc_data, aes(x = tissue, y = perc_sparc_pos_oligos)) + geom_boxplot() +
geom_jitter(width = 0.2, aes(colour = sample_id)) +
theme_bw() + ylab("SPARC+OLIG2+RBFOX1-HOECHST+ (%)") + xlab("Tissue")
```
```{r}
ggplot(sparc_data, aes(x = tissue, y = perc_sparc_pos_oligos)) + geom_boxplot() +
geom_jitter(width = 0.2, aes(colour = donor_id)) +
theme_bw() + ylab("SPARC+OLIG2+RBFOX1-HOECHST+ (%)") + xlab("Tissue")
```
```{r}
ggplot(sparc_data, aes(x = tissue, y = total_sparc_oligos_perc)) + geom_boxplot() +
geom_jitter(width = 0.2, aes(colour = donor_id)) +
theme_bw() + ylab("SPARC+OLIG2+HOECHST+ (%)") + xlab("Tissue")
```
```{r}
ggplot(sparc_data, aes(x = tissue, y = perc_sparc_pos_rbfox1_pos_olig2_pos)) +
geom_boxplot() +
geom_jitter(width = 0.2, aes(colour = donor_id)) +
theme_bw() + ylab("SPARC+OLIG2+RBFOX1+HOECHST+ (%)") + xlab("Tissue")
```
```{r}
sparc_data$Num_Detections<- as.numeric(sparc_data$Num_Detections)
mod <- glmer( Num_SPARC_OLIG2 ~
tissue +
Num_Detections +
(1 | u_id)+
(1 | u_fov) ,
data = sparc_data, family = poisson(link = "log"))
summary(mod)
```
```{r}
lm_mod <- lm( Num_SPARC_OLIG2 ~
tissue +
Num_Detections,
data = sparc_data)
summary(lm_mod)
```
```{r}
sparc_data$cube_rt_perc_sparc_pos_oligos <- pracma::nthroot(sparc_data$perc_sparc_pos_oligos, 3)
sparc_data$log10_perc_sparc_pos_oligos <- log10(sparc_data$perc_sparc_pos_oligos)
lm_mod <- lm( perc_sparc_pos_oligos ~
tissue +
age_group * sex,
data = sparc_data)
summary(lm_mod)
lm_mod <- lm( cube_rt_perc_sparc_pos_oligos ~
tissue +
age_group * sex,
data = sparc_data)
summary(lm_mod)
AIC(lm_mod)
```
```{r}
red_mod <- lm( cube_rt_perc_sparc_pos_oligos ~
tissue,
data = sparc_data)
AIC(red_mod)
```
```{r}
hist(resid(lm_mod))
shapiro.test(resid(lm_mod))
```
Is the difference between CB and CSC statistically significant?
```{r}
sparc_data$tissue_2 <- ifelse(sparc_data$tissue == "CSC", "A_CSC",
paste(sparc_data$tissue) )
lm_mod <- lm( cube_rt_perc_sparc_pos_oligos ~
tissue_2 +
age_group * sex,
data = sparc_data)
summary(lm_mod)
```
```{r}
sessionInfo()
```