Affymetrix Experimental Design
Contact Anne-Marie Girard to schedule a consultation about experimental design. Below are a number of commonly asked questions to help you begin planning your experiment. Please consult with us to let us know the time frame of your experiment and to give you an idea of our current workload.
- What question(s) are you trying to answer with microarrays? If you have more than one question, you may need more than one experiment. A complex problem may be best addressed with more than one experiment.
- Does this experiment fit into one you have already run? Will you be comparing between these experiments, or with future experiments? The same RNA isolation methods, the same amount of starting material and method for labeling must be used. If we know it needs to fit in with another experiment, we can make sure the same conditions on our end are used.
- Can you suggest any good references in preparing for a microarray experiment?
- Nature Reviews Genetics (2004): Expression Profiling - Best Practices for Data Generation and Interpretation in Clinical Trials. It gives a good introduction into the best practices for performing a microarray study.
- Nature Reviews Drug Discovery (2002): The Use and Analysis of Microarray Data
- Nature Methods had several articles comparing types of microarrays.
- Affymetrix Data Analysis Fundamentals Guide describes experimental design, assessing data quality, some statistics and data interpretation, along with their annotation mining tools.
- What experimental variables may affect my microarray experiments?
- Sample collection and RNA isolation. All variables that you do not want to study should remain the same. In other words, samples should be collected at the same time of day, the same amount of time after watering or feeding. Biological samples kept near the air conditioning will show different gene expression patterns compared to those further away. Greenhouse grown plants grown in the winter will show different patterns than those grown in the summer. A growth chamber will provide more consistent conditions if you want to run experiments year round.
- RNA concentration and quality should be checked on a reliable spec. We have a Nanodrop Spectrophotometer available to read low concentrations. Make sure the amount of starting material is the same for every chip in the experiment.
- Using different technicians, equipment and a variety of reagent lots may also affect your results.
- How will pooling samples affect my data? Here are a few references that discuss pooling.
- Affymetrix Technical Report on Pooling
- Kerr, M.K., Design considerations for efficient and effective microarray studies. Biometrics 59, 822-828 (2003).
- Kendziorski, C.M., Zhang, Y., Lan, H., and Attie, A.D., The efficiency of pooling mRNA in microarray experiments. Biostatistics 4, 465-477 (2003).
- Peng, X., Wood, C.L., Blalock, E.M., Chen, K.C., Landfield, P.W., and Stromberg, A.J., Statistical implications of pooling RNA samples for microarray experiments. BMC Bioinformatics, 4, 26 (2003).
- C. Kendziorski, R. A. Irizarry, K.-S. Chen, J. D. Haag and M. N. Gould, On the utility of pooling biological samples in microarray experiments. PNAS, 102, 4252-4257.
- How many replicates should I use in my experiment? To be able to use standard T-tests, you should have at least 3 replicates. ANOVA would need 5 replicates. The more replicates you have the lower the False Discovery Rate (FDR) ( Wolfinger et al 2001.) Fewer replicates will identify a smaller percentage of the differentially expressed genes. BMC has a website which shows the influence of number of replicates on the FDR .
- How do researchers verify their microarray data? What is available through the CGRB Core Labs? RT-PCR, Northerns and quantitative PCR are a few of the methods that are used to verify genes from microarray experiments. The CGRB Core Labs have the ABI 7000 Sequence Detection System for performing quantitative PCR.