home
information
experimental design: variation
It is important to try and develop a model that effectively deals with the many sources of variance that can be introduced. With a microarray experiment, there are four major contributors to variance:
The genes or probes which are spotted:
- Each probe spotted on the array may differ in length or GC content thus affecting the hybridisation efficiency and kinetics
- Each corresponding cDNA or cRNA that is labelled and hybridised to the array will differ in length, labelling efficiency and abundance
The arrays themselves:
- Each array is made separately even if it was made in a single batch and thus may exhibit differences in spot quality
- Each array is printed onto a separate piece of glass or silicon
- Often times, a project may utilise arrays from multiple batches/lots
The dyes being used:
- Each of the dyes used differ in photostability, brightness, detectability and incorporation efficiency
- A large experiment may require use of dyes from different batches/lots
The treatments involved:
- Each treatment is performed individually and will have its own effects
- Treatments may be done at different times
- Treatments may even be done by different people