1.Overview of the data sets.Rmd

---
title: "Overview of the data sets"
author: "Bai Ling"
date: "`r format(Sys.Date(), format = '%d %B %Y')`"
output: 
  html_document:
    theme: cerulean
    css: style.css
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, fig.align = 'center')
options(stringsAsFactors = FALSE)
setwd("/Users/apple/Documents/GitHub/EWAS/")
```

## Contents
***
- [1. Tissue distribution of EWAS data sets](#tissue)
- [2. QQ-plot of p-values](#qqplot)
<br>
<br>
<br>

#### Library R packages and basic settings

```{r, message=FALSE, warning=FALSE}
library(tidyverse)
library(readxl)
library(ggpubr)
options(stringsAsFactors = FALSE)
```

### 1. Tissue distribution of EWAS data sets {#tissue}
***
The 61 EWAS data sets' tissue sources and sample sizes are displayed in the following plot, and the majority data sets are form blood.
```{r}
sampleinfo <- read.csv("Data/DatasetInfo.csv", header = TRUE)
sampleinfo <- arrange(sampleinfo, desc(Size))
sampleinfo$Tissue <- factor(sampleinfo$Tissue, levels = unique(sampleinfo$Tissue))
sampleinfo <- arrange(sampleinfo, Tissue, desc(Size))
sampleinfo$New_ID <- factor(sampleinfo$New_ID, levels = unique(sampleinfo$New_ID))
my_palette <- c("#1B9E77", "#D95F02", "#7570B3", "#E7298A", "#E6AB02",  "#A6761D", "#A6CEE3", "#1F78B4", "#B2DF8A", "#FFFF00", "#FB9A99", "#E31A1C", "#949494", "#FDBF6F", "#FF7F00", "#CAB2D6", "#2F4F4F", "#7FFFD4", "#FFE4E1", "#00E5EE")
fs1 <- ggplot(data = sampleinfo, aes(x = New_ID, y = Size, group = Tissue)) +
  geom_bar(aes(fill = Tissue), stat = "identity", width = 0.5) + 
  scale_fill_manual(values = my_palette) +
  theme_classic() + 
  theme(axis.text.x = element_text(angle = 75, size = 5, hjust = 0, vjust = 0.1), 
        axis.title.x = element_blank(),
        legend.position = "top",
        legend.key.size = unit(0.3, "cm"),
        title = element_text(size = 10),
        axis.text.y = element_text(size = 8),
        legend.text = element_text(size = 8))
fs1
```

```{r, include=FALSE}
ggsave("Figures/FigureS1.pdf", plot = fs1, width = 5, height = 4, useDingbats = FALSE)  
```

### 2. QQ-plot of p-values {#qqplot}
***
The genomic inflation of association p-values before and after surrogate variable analysis are shown as QQ-plot. As the source code about QQ-plot is provided in code folder, file qqplot.RData is not upload here beacuse the file size is too large.
```{r, results='hide', fig.width=6.69, fig.height=8.86}
load("Data/qqplot.RData")
myplots <- list()
for(i in 1:length(myplot)){
  myplots[[i]] <- myplot[[i]] + ggtitle(names(myplot)[i]) +
    theme(axis.title = element_text(size = 6), axis.text = element_text(size = 6), plot.title = element_text(hjust = 0.5, size = 8))
}
plots <- ggarrange(plotlist = myplots, ncol = 4, nrow = 4, common.legend = TRUE)
# show the example of 16 datasets
plots$`1`
```

```{r, include=FALSE}
pdf("Figures/FigureS2.pdf", width = 6.69, height = 8)
plots
dev.off()
```