Module 4 of E0 259, Data Analytics, August 2019

Effects of Smoking

Lectures

Data set

Assignment 4

Due: 23:55 hrs, Wednesday 30 October 2019. Discussion is encouraged. But write your own code and submit your own work. Please comply with the ethics policy.

Your goal is to identify genes which respond differently to smoke in men vs. women (Smoking Status X Gender model vs. Smoking Status + Gender null model)

1. Use the 2-way ANOVA framework to generate p-values for each row of the data set

2. Draw the histogram of p-values

3. See if a better (than n estimate for n₀ is derivable from this histogram; justify your estimate.

4. Use an FDR cut-off of 0.05 to shortlist rows.

5. Create a shortlist of gene symbols from these rows.

6. Intersect with the following gene lists: Xenobiotic metabolism, Free Radical Response, DNA Repair, Natural Killer Cell Cytotoxicity. (Click on links to get the text files containing the lists.)

7. Report intersection counts for each list, split into four groups; going down in women smokers vs. nonsmokers / going up in women smokers vs. nonsmokers. Do the same for men.