01-preprocessing.qmd

---
title: "Untitled"
format: html
---

## Setup

Remember, we started with a `SummarizedExperiment` object and now we have to transform it into an object that is usable with `DESeq2`. 

```{r}
#| warning: false
#| message: false
library(SummarizedExperiment)
library(tidySummarizedExperiment)
library(ggplot2)
library(DESeq2)

GSE96870 <- readRDS("data/GSE96870_se.rds")
```


## Plot

```{r norm-library}
#| echo: false
library(patchwork)

expression_plot <- 
  GSE96870 |>
    filter(.feature == "Asl") |>
#    filter(time %in% c("Day0", "Day8")) |>
    ggplot(aes(x = Label, y = counts, color = time)) +
    geom_point() + theme_minimal() +
    theme(axis.text.x = element_blank()) + 
    ggtitle("Asl Expression count") 

count_plot <- 
  GSE96870 |>
#    filter(time %in% c("Day0", "Day8")) |>
    group_by(Label) |>
    summarise(total_counts=sum(counts)) |>
    mutate(time = colData(GSE96870)[["time"]]) |>
    ggplot() +
    aes(x=Label, y=total_counts, fill=time) +
    geom_bar(stat="identity") + theme_minimal() +
    theme(axis.text.x = element_text(angle = 90)) +
    ggtitle("Library Size for each Sample") 

expression_plot/ count_plot 
```

## Normalizing

```{r}
GSE96870_filtered <- GSE96870[rowSums(assay(GSE96870, "counts")) > 5,]
GSE96870_deseq <- DESeq2::DESeqDataSet(GSE96870_filtered, design = ~ sex + time)
GSE96870_normalized <-  DESeq2::estimateSizeFactors(GSE96870_deseq)
GSE96870_normalized |>
  select(.feature, .sample, counts, sizeFactor)
```

```{r}
expression_normalized <- GSE96870_normalized |>
    filter(.feature == "Asl") |>
    mutate(scaled = counts / sizeFactor) |>
    ggplot(aes(x = Label, y = scaled, color = time)) +
    geom_point() + theme_minimal() + 
    theme(axis.text.x = element_blank()) + 
    ggtitle("Asl Expression count \nnormalized by library size")

(expression_plot + theme(legend.position = "none")) + expression_normalized
```