R Tutorial: Differential expression analysis

---
Welcome to differential expression analysis with limma. I'm John Blischak, and I'll be your instructor for this course.

You will learn how to analyze data generated by functional genomics experiments, so let's review the terminology I will use throughout the course. Imagine a hypothetical experiment in which the researcher has treated cells with two different drugs, A and B. The drug treatment is the variable of interest, and is an example of a phenotype.

Each sample is processed by a high-throughput assay to measure thousands of genes. These are the features. In this course the features will always be genes, but in other experiments they could be proteins or some other molecular feature of a cell.

For each feature, the assay produces a value that is a proxy for the relative abundance of that feature. For genes, this is the number of RNA transcripts that are expressed. Thus I will refer to the measurements as expression levels.

The statistical models you will build in this course will test for differences in these measurements between samples with different phenotypes. If a feature has a higher expression level for one group relative to the other, this is called upregulated. Conversely, a lower expression level is called down-regulated.

The overall goal is to identify the genes that are associated with a phenotype of interest. Some examples of differential expression studies would be identifying all the genes associated with a response to a stimulus like a drug, a developmental process, or a genetic mutation.

Now why bother testing thousands of genes, known as a genome-wide differential expression analysis, when it would be easier to focus on measuring only a handful of genes? The first motivation is novelty. You may discover that some additional genes play an unexpected role in the process you are studying. Second is context. Interpreting the relevance of any one gene is easier when you can compare it to the behavior of all other genes. Third is to gain a systems-level understanding of the process. By measuring all genes simultaneously, you can identify higher-order systems like pathways that you would miss with a more targeted approach.

There are many steps to complete an experiment, from designing the initial study to compiling the results. In this course, the main focus will be on creating linear models to test scientific hypotheses, but you will also learn the basics of exploring the data and interpreting the test results. Thus in general this course is agnostic to the type of data collected. To learn more about how to process the data generated by a specific technique, check out the other Bioconductor courses on DataCamp.

Now for some caveats to keep in mind. First, the measurements recorded by these genomic technologies are relative, not absolute. Thus comparing the measurements across samples is valid, but it's not possible to directly convert these measurements to the total number of molecules. Second, study design is very important. All experimental techniques introduce some technical noise, but the statistical methods for removing this noise are only valid for properly designed studies. You'll learn about proper study design later in the course.

Now let's test your comprehension.

#R #RTutorial #DataCamp #Analysis #limma #Differential #expression #data

[johirislam]

I want have some other quaries regarding DEG analysis.I want to compare
two datasets differentially expressed gene, how can i do that.For
example one data set contain 108 DEG and the other contain 70 so i want
to see the common gene between this two dataset.So how can i do that and
how can i make the vaan diagram between them.Moreover i saw some GEO
dataset there are some file format tsv and txt.Son in that case how can i
analyse that kind of file.Plz solve this two problem to me.

[Jonpaulim]

great video can I ask a question? do you know if LIMMA Voom is suitable for microarray analysis?

[lacomp]

🖌️... ..








different equations?