Activity 1

The analysis of gene expression will allow us to know the different levels of expression between various groups of interest, for example between sick and healthy people. This information is very interesting because it helps us to better understand the molecular mechanisms of a disease, it will facilitate the detection of biomarkers, …..

We have prepared the following activity to familiarise ourselves with these common scenarios in gene expression analysis. We will use the Babelomics suite, which contains several web tools for omics data analysis.

We are interested in detecting therapeutic targets from the evaluation of various pharmacological treatments in patients with a given disease.

• We have a total of 49 patients and 5 types of treatment.
• We randomly gave each patient a single treatment out of the 5 available: placebo (code 1) and several treatments (codes: 3,5,7 and 9).
• So each treatment was allocated to 10 individuals, except for treatment 5 which was allocated to 9 individuals.
• We then drew blood to determine the expression levels of all the genes using microarrays. After hybridisation we have the following data.

We will carry out a bionformatics analysis that will allow us to cover the objective of the activity.

• We access the Babelomics web tool from the Firefox or Chrome browsers: http://courses.babelomics.org/
• We register in the tool

Exercise 1. We will do this first exercise together.

1. We go to the ‘Expression / Arrays /Differential Expression’ menu.
2. We run the online example and select two groups to compare: patients with placebo (code 1) and patients with treatment 3.
3. Run the job. What do you think of the results?
4. How many genes are differentially expressed between the two treatments?
5. Download the file with the genes that are differentially expressed (significant).
6. How do you interpret the heatmap that appears in the results?
7. Which gene shows the greatest change in expression between treatments? Look for detailed information on its function at Ensembl:http://www.ensembl.org/index.html

Exercise 2. : We continue with the previous study. Now we would be interested to know if there are differentially expressed genes between placebo and treatment 5. Also between placebo and 7, and finally: placebo vs. treatment 9.

Some questions:

1. How many genes are differentially expressed between the two treatments?
2. Download the file with the differentially expressed (significant) genes.
3. How do you interpret the heatmap that appears in the results?
4. Which gene shows the greatest change in expression between treatments? Find detailed information about its function at Ensembl:http://www.ensembl.org/index.html

Exercise 3. Common and specific biomarkers.

• Finally, are there any common genes in the results of these three comparisons: 1vs3, 1vs7 and 1vs9? Use the Venn diagrams provided by the Venny tool. Try changing the format of the graphical representation with the options provided by this tool.
• How would you interpret the results obtained? Do all treatments show the same results in the differential expression analysis? Do you think it would be important to study the function of these genes of interest?

Results:

• Significant gene intersection analysis: 92common_biomarkers.txt
• Significant genes in each placebo vs. treatment comparison: placebo vs. t3, placebo vs.t5, placebo vs. t7, placebo vs. t9.

Exercise 4: functional profiling

• Functionally characterise the group of genes common to all comparisons using an overrepresentation analysis of biological functions (we will start with GO-slim biological processes). We will then characterise these same genes with a protein-protein interaction analysis.
  • Let's go to Panther webtool (http://www.pantherdb.org/)
  • Let's go to STRING webtool (https://string-db.org/)
• Repeat the above approach for genes specific to one of the treatments.