Genomics data analysis: the goal is to identify regions of interest through different methodologies.

Differential analysis aims to identify differences in expression between two biological conditions. Usually it’s tested against the null hypothesis of no difference through a statistical test. In order to do that, we need:

Then, we define a statistic, derive the distribution of statistics under H0, and calculate a p-value. After choosing a alpha, we decide wether to accept or reject the null-hypothesis. Likely, we will commit type I and II errors. In fact, because usually we are making thousands of tests, we should apply some multiple-test correction.


Assessing the quality of differential expression analysis

We will evaluate three statistical frameworks to model gene expression:

What do we want to know:

How are we going to evaluate the models: