delta-beta-calc.Rmc

---
title: "Compare DEG tools"
author: "Mark Ziemann"
date: "`r Sys.Date()`"
output:
  html_document:
    toc: true
theme: cosmo
---


## Introduction

In this tutorial we will demonstrate how to calculate delta-beta values in an epigenetics study starting with a matrix of beta values which contains disease and normal samples.
The general approach is to filter these samples (columns) into separate matrices and then calculate the mean beta value for each probe then calculate the delta-beta.

## Fake data generation

The first part is to create a matrix of data. 
In this case the matrix will be 10 rows (genes) and 10 columns (samples).

```{r,make_data}

vec <- rnorm(n = 100,sd = 1,mean = 0)
mx <- matrix(data = vec,nrow = 10)
colnames(mx) <- paste("sample",1:10)
rownames(mx) <- paste("probe",1:10)

mx[1:6,1:5]

```

Now need to decide which samples are control and which are disease.
This will be coded in your sample sheet but here I'll just select them by random.

```{r,samplegroups}

my_disease_samples <- sample(x = colnames(mx),size = 5,replace=FALSE)
my_disease_samples

my_ctrl_samples <- setdiff(colnames(mx),my_disease_samples)
my_ctrl_samples

```

## Column filtering

Now need to filter the matrix based on the column names.

```{r,colselect}

mx_disease <- mx[, colnames(mx) %in% my_disease_samples]
dim(mx_disease)

mx_ctrl <- mx[, colnames(mx) %in% my_ctrl_samples]
dim(mx_ctrl)

```

## Calculate beta and delta-beta values

Next, calculate the beta values for the two groups and find the difference

```{r,betavals}

betavals_disease <- rowMeans(mx_disease)
betavals_ctrl <- rowMeans(mx_ctrl)

deltabetavals <- betavals_disease - betavals_ctrl

head(deltabetavals)

```