FGSR Experiment Design

The design of your microarray experiment directly influences the precision of your results. Because of the large-scale, sensitive nature of microarrays, the experimenter must control for as many differences between organisms as possible. This page contains guidelines that we have found helpful in experimental design, but every project is different. For help in designing an experiment to suit your requirements, please contact us.

Replicates: The FGSR strongly advises all investigators to include at least three biological replicates for each experimental condition, including control. By including replicates in your experiment, much more powerful techniques can be applied to your analysis, resulting in more fruitful data. To maximize the benefit of these extra samples, you should choose your replicates carefully. They should be different organisms, but everything else should be the same among groups. The animals should be the same sex, and should be littermates. They should be raised in the same conditions and given the same treatments at the same time. They should be as genetically similar as you can make them. Obviously, it can be difficult in a large experiment to meet every single one of these conditions. Care should be taken so that your replicates are as similar as possible.

We stress the use of biological replicates. Technical replicates (ie, hybridizing the same sample on three different arrays) will only assess variations in manufacturing, dye incorporation, pipetting, etc. These variations tend to be much smaller than the real variations between organisms, and they can be corrected through data normalization. Therefore, we feel that the most value is achieved through the proper use of biological replicates.

Pools: Pooling samples can be an excellent way to get answers in situations where your RNA amounts are low. However, great care should be taken when setting up the pools. When choosing which samples to pool, follow the same guidelines as you would for choosing replicates. Keep your pool size consistent - comparing a two-sample pool with a four-sample pool is not as useful as comparing two three-sample pools. Finally, you still need replicates! Four pools of three animals only yields four data points, not twelve. Even if your samples are pooled, you still need replicates to be able to apply statistical tests.

Patient Samples: We are happy to run patient samples, but the investigator must keep several things in mind:

  1. We are not a HIPAA-compliant facility. All human samples must be completely de-identified before we see them - the tubes should only have the FGSR ids on them, and you should not include any identifying information on the sample submission sheet.
  2. We are not a clinical laboratory. All of our procedures are approved for research only, not for diagnostic purposes
  3. Humans have a very deep genetic pool - much deeper than any laboratory animals. There will be, on average, more variations in human samples than mouse, and they are harder to control. For instance, human replicate samples are rarely "litter mates". When possible, we advise the use of more than three replicates, and sample pools can also help