---
title: 'WGCNA: data preprocessing'
output:
    html_document:
      number_sections: true
      toc: true
      theme: cosmo
      highlight: tango
---

```{r setup, include=FALSE, warning=FALSE, message=FALSE}
knitr::opts_chunk$set(
  echo = ECHO
)
```

```{r}
str(opt)
```

# Import data

Each row represents a gene and each column represents a sample.

```{r}
expression_data = read.csv('EXPRESSION_DATA', header = TRUE, row.names = 1)
```

Display the first 100 genes.

```{r}
datatable(head(expression_data, 100), style="bootstrap", filter = 'top',
          class="table-condensed", options = list(dom = 'tp', scrollX = TRUE))
```

Transpose expression data matrix so that each row represents a sample and each column represents a gene.

```{r}
expression_data = as.data.frame(t(expression_data))
```

# Checking data

Checking data for excessive missing values and identification of outlier microarray samples.

```{r}
gsg = goodSamplesGenes(expression_data, verbose = 3)
if (!gsg$allOK) {
  # Optionally, print the gene and sample names that were removed:
  if (sum(!gsg$goodGenes)>0)
    printFlush(paste("Removing genes:", paste(names(expression_data)[!gsg$goodGenes], collapse = ", ")));
  if (sum(!gsg$goodSamples)>0)
    printFlush(paste("Removing samples:", paste(rownames(expression_data)[!gsg$goodSamples], collapse = ", ")));
  # Remove the offending genes and samples from the data:
  expression_data = expression_data[gsg$goodSamples, gsg$goodGenes]
} else {
  print('all genes are OK!')
}
```

# Clustering samples

If there are any outliers, choose a height cut that will remove the offending sample. Remember this number since you will need this number in further analysis.

```{r fig.align='center'}
sampleTree = hclust(dist(expression_data), method = "average");
plot(sampleTree, main = "Sample clustering to detect outliers", sub="", xlab="",
     cex.axis = 1, cex.main = 1, cex = 0.5)
```


```{r echo=FALSE}
rm("opt")
save(list=ls(all.names = TRUE), file='PREPROCESSING_WORKSPACE')
```