Statistical Analysis of Microarray Data by Hanne Jarmer
Statistical Analysis of Microarray Data by Hanne Jarmer
Induction of adherence by sub-lethal alcohol concentrations
Induced adherence - Listeria monocytogenes adheres better at sub-lethal alcohol concentrations Microarray experiments on: 1. Wild type +/- alcohol 2. Mutant +/- alcohol 4 biological replicates of each
The microarray data (wild type)
The microarray data (mutant)
Why and what do we test? - We test to extract significant genes Density Fold change: > 2 Intensity Fold change: ~1
The t-test The t statistic is based on the sample mean and variance t
The P-value Definition: The possibility of getting the observed difference by coincidence
Correction for multiple testing Each time we test, there is a certain possibility, that the observed difference is in fact a coincidence when H 0 is TRUE Unacceptable many false positives
Correction for multiple testing Bonferroni: Confidence level of 99% 0. 01 P≤ N Benjamini-Hochberg: P≤ i N 0. 01 N = number of genes i = number of accepted genes
Volcano plot P-value log 2 fold change (M)
The 2 way ANOVA 2 Interaction 3 wildtype +alcohol 1 mutant +alcohol 3 wildtype 3 mutant 2 1
What would be significant? Intensity 1. +/- alcohol wt mutant alcohol both 2. +/- mutation wt mutant alcohol both 3. +/- both wt mutant alcohol both
Acknowledgements Anne Lise Gravesen, KVL Growth experiments Torsten Hain, Institut für Medizinsiche Mikrobiologie Microarray experiments
Correction for multiple testing Bonferroni: Confidence level of 99% 0. 01 P≤ N Benjamini-Hochberg: P≤ i N 0. 01 N = number of genes i = number of accepted genes
- Slides: 15