Peter from Macquarie University introduces ChIP-seq, a technique that identifies where proteins bind DNA in the genome. The course uses a prostate cancer dataset with five primary tumors and three treatment-resistant samples, all immunoprecipitated for the androgen receptor. You'll learn R functions like readAlignments, coverage, and GRanges to work with BAM and BED files.
Chapters
Hi and welcome to Chip SEQ Workflows in R. My name is Peter. I'm a statistician at Macquarie University and will be your instructor for this course. I'm assuming that you have some experience with R and have walked through one of the introductory BioConductor courses on datacamp. In this course, you will learn how to process and analyze ChIP seq data in R. Before we get into the meat of the course, let us briefly talk about what Chip SEQ is and how it is used. The question at the center of a CHIP SEQ experiment How do cells know what to do? Your body contains several trillion cells, each carrying the same genome, yet these cells do different jobs. Some are neurons in your brain, some form your skin, and some flow through your body as blood cells. What is it that makes these cells different from each other? The function of a cell is in large part determined by the genes that it expresses. Genes encoded in the DNA that makes up the genome are transcribed into RNA and then translated into proteins. The resulting proteins are responsible for carrying out the cell's function. Inside each cell, a complex machinery of proteins is responsible for ensuring that the right genes are translated into proteins. Inhibitors are proteins that bind to DNA to deactivate specific genes. Such inhibitors have to be removed through the interaction with other proteins before genes can be expressed. A complex of activating transcription factors can then bind to the DNA, allowing gene expression to proceed through this process. The regulatory machinery ensures that a cell correctly performs its role. If these regulators are disrupted, cells can get out of control, causing a variety of diseases, including cancer. Chip, short for Chromatin immunoprecipitation, is a technique that can be used to extract specific proteins together with any part of the genome they were bound to from a cell. We can then use the DNA sequences attached to the proteins to infer the sites across the genome that they interact with by identifying regions of the genome that are over represented in the sequencing data. By comparing these binding sites between different individuals, for example Healthy volunteers and cancer patients, we can uncover the mechanisms that are responsible for the differences between them. Throughout this course, you will explore one particular CHIP SEQ dataset. This data consists of samples taken from patients with prostate cancer and fall into two groups. The first group consists of five primary tumor samples, while the second contains three samples from tumors that have become treatment resistant. Chromatin immunoprecipitation targeting the androgen receptor was carried out for all samples. The androgen receptor is activated by binding testosterone and is then capable of binding to specific parts of the genome to regulate gene expression. Before we move on to some exercises, let's review a few functions used in R to interrogate sequencing data. Mapped sequence reads are typically stored in BAM files. You'll learn more about this type of file in chapter two. Using the Read alignments function, you can load data from BAM files. Once loaded, information about the chromosome reads have been mapped to is available via a call to the seqnames function, and their location on that chromosome can be accessed through the functions start and end, respectively. We often want to know the location of mapped reads, but also how many reads cover any given position in the genome. This information can be computed with the coverage function. The main units of interest in the analysis of a CHIP SEQ experiment are peakalls that highlight regions of the genome with a high concentration of reads. These peak calls are typically stored in bad files. We look at what these files look like in more detail later. Each peak is associated with a score, which quantifies the strength of this particular peak. Peak calls can be loaded with the import BAT function. Coordinates of peaks can then be obtained by calling the Chrome and Ranges functions. The score function provides access to the peak scores. Time to put this into practice.
