Experimental design of DNA chip and microarray analyses 


A short introduction to experimental design (UCL, London)

Designing a microarray experiment (Microarray Facility, Genomics Lab, UMC, Utrecht)

Experimental design has also been discussed during the fourth MGC-course Analysis of microarray gene expression data


Brief overview:

Before you start the experiment:

- Clearly determine your research goal.

- Decide how you will perform your data analysis

- Determine which samples you want to use

    In vivo:

    Human tissue sample (tissue 1 vs. tissue 2; normal vs. tumor; subject 1 vs. subject 2, etc.)

    Animal model (normal vs. mutant; background 1 vs. background 2)

    In vitro:

    Cell line (normal vs. tumor or mutant; background 1 vs. background 2)

- Determine which treatment or sample times you want to use

    Treatment vs. control / Disease vs. control

    Multifactorial design (e.g., genotype and treatment)

    Time course

- Estimate the number of replicates needed

Replicates are needed to increase the precision of the measurement and for error estimation.

True replicates: multiple samples taken to account for biological variability.

Repeated measurements: 

- multiple spots per gene per microarray or DNA chip

- same mRNA samples used on multiple microarrays or DNA chips

- For microarrays: label each sample with both dyes!

Discuss your experimental design with a statistician!


Sources of variability:

Subject variability due to genetic and/or environmental differences

Sample variability due to differences in tissue isolation, handling, and storage, differences in cell culture conditions, handling, and storage.

Alternative information and sites about gene expression and microarray analysis can be found through Y.F. Leung's microarray site