2025-03-21

Exercise summary

1. Read the following two RDS files available from Datasets folders:

• Differential_expression_deseq2_pc3.rds: dataset including gene expression results of PC3 cell line replicates
• Differential_expression_deseq2_vcap.rds: dataset including gene expression results of VCAP cell line replicates

How many genes are in each table?

---

2. For each dataset, select differentially expressed genes, defined as genes with:

• padj ≤ 0.01 and |log2FoldChange| ≥ 1.

Create a new column diff based on the sign of log2FoldChange:

• If log2FoldChange > 0 = up
• If log2FoldChange ≤ 0 = down

How many differentially expressed genes are there per cell line model? How many are up-regulated and down-regulated?

Summarize the results in a table.

---

3. Create a Venn diagram to show the intersection of differentially expressed genes between VCAP and PC3.
   Ensure that the cell line names are labeled for each category in the plot.

---

4. Create an UpSet plot to visualize the intersection of the following gene sets:

• up-regulated in PC3
• up-regulated in VCAP
• down-regulated in PC3
• down-regulated in VCAP

---

5. Select genes that are up-regulated in both PC3 and VCAP cell models. Add the correspondent log2FoldChange values from both PC3 and VCAP experiments. How many such genes are there?

Now, create a scatter plot, using:

• log2FoldChange values from both datasets
• baseMean as the size of the points.

Evaluate the correlation between log2FoldChange values.

---

6. Select genes that are down-regulated in both PC3 and VCAP cell models. Add the correspondent log2FoldChange values from both PC3 and VCAP experiments. How many such genes are there?

Now, create a scatter plot, using:

• log2FoldChange values from both datasets
• baseMean as the size of the points.

Evaluate the correlation between log2FoldChange values.

---

7. Arrange the scatter plots of up-regulated and down-regulated genes together in a single layout for better visualization and comparison.

---

8. Create a function to classify genes based on their log2FoldChange values:

• LU (Lowly upregulated): log2FoldChange > 0 and < 2
• MU (Moderately upregulated): log2FoldChange between 2 and 5
• HU (Highly upregulated): log2FoldChange > 5
• LD (Lowly downregulated): log2FoldChange < 0 and > -2
• MD (Moderately downregulated): log2FoldChange between -2 and -5
• HD (Highly downregulated): log2FoldChange < -5

---

9. Create a unified dataframe all, by concatenating the differentially expressed gene dataframes from PC3 and VCAP.

Before merging, add a new column to each dataframe to indicate the cell model (PC3 or VCAP) for proper distinction.

---

10. Apply the function created in Exercise 8 to all values in the log2FoldChange column of the all dataframe:

• Store the results in a new column called de_category.

---

11. Create a barplot to visualize the distribution of genes across categories de_category.

---

12. Now, assess whether gene length impacts the category of differentially expressed genes.

13. Create a TxDb object from the GTF file:
    File: gencode.v28lift37.basic.annotation.gtf.gz (located in the Datasets folder).
    

14. Extract genes that are present in the all dataframe.
    

15. Compute gene length for these genes.
    

16. Add the gene length as a new column in the all dataframe.

---

13. Create a boxplot to visualize gene lengths across cell line models (PC3 and VCAP) within each differential expression category de_category.
    

Add Wilcoxon test results to assess whether there is a statistically significant difference in gene length between PC3 and VCAP within each de_category.

---