Gene Expression Profiling Submission Requirements

How many samples should you include in your experiment?

Since the Illumina bead arrays measure each transcript an average 30 times in separate locations on the array, the sample-to-sample variance for duplicates is small (r2 >0.996). Therefore, we recommend against running replicates of the same sample, and instead we recommend that you increase the number of biological samples in order to identify the important significant differences between groups, and not just random differences between individual samples. For a pilot study, we recommend starting with 4 vs. 4 samples with a single 24k array, or for a large study, examine 6 vs. 6 samples with the 48k array (2 chips). It is important to note that with a higher number of samples the real differences between samples will be more significant, and random effects will be less so, resulting in less noise to wade through.

Sample submission guidelines for RNA expression experiments

• RNA extraction

  Please prepare high quality total RNA. During RNA purification, you should work quickly and add the homogenization buffer to your samples without allowing the samples to thaw before rapidly homogenizing (we use a Q-BioGene bead-shaker with tissue lysing beads). For RNA extraction we have had excellent success with the Qiagen RNAeasy mini kit (Qiagen catalog no. 74104, 50 preps).

  The RNA should be suspended in pure Rnase free water with absolutely no SDS or any other detergent in the solution and free of any other organics.

• RNA concentration

  For your experiment, ensure that your samples have equal starting quantities of RNA. We recommend checking total RNA concentration using the Quant-it Ribogreen RNA kit (Invitrogen catalog no. R11490).

  The recommended quantity of RNA for expression analysis is 250 ng of total RNA in 16 ul volume of pure Rnase free water in a 96 well plate format.

• RNA BioAnalyzer quality check

  We strongly recommend sending the RNA samples to the CMMT BioAnalyzer Core Facility for analysis on the Agilent 2100 BioAnalyzer. The cost is minimal compared to the expression analysis cost. The BioAnalyzer can predict whether a sample will perform properly and identify samples that are not of sufficient quality.

• Sample drop off

  Please label samples well and place into a pre-labeled freezer box, transport on dry ice, and store at -80°C.

What you will get back from us

We will send you the raw data (or upload it to our accessible network site), and send you a program to analyze the data with (BeadStudio), and the software manual, along with our "simplified manual" of step-by-step instructions on how to visualize the data and export the important data into excel.

What your data will look like (and what you can do with it)

You will receive expression signal data for 24,000 to 46,100 transcripts for each individual sample and the standard deviation for each individual bead's measure (average 30 beads/transcript). In addition, we will show you how to sort your individual samples into groups, in order to examine the data for each group (+/- SD) to quickly identify significant differences between groups.

In addition:

• For each transcript you will receive up-to-date information about the gene, including, gene symbol, accession ID, synonyms, gene definitions, and GO functions and processes.

• You will be able to easily graph your data in a scatter plot to visually identify outliers and show X-fold changes.

• You will be able to highlight genes by GO code function on scatter plots (i.e. show only genes related to lipoprotein metabolism, and determine whether these show a similar trend in your experimental group vs. controls).

• You will be able to cluster by sample (automatically group similar samples by expression profiles).

• You will be able to cluster by gene (automatically cluster similar gene expression patterns).

• For each transcript, you will receive a score for "detectability" because some genes are not expressed in all tissues, and if they do not reach a minimum threshold they will be excluded from downstream analyses if you want (recommended).

• In addition, the data is exportable to 3rd party gene expression analysis software programs.

Return to gene expression profiling services.