Experimental Design and Controls

Define probe v. target

Label RNA v. cDNA (impact on spotted material)

Cy series v. Alexa series dyes

Spots made of oligos v. cDNA

Affymetric chips v. printed glass chips

Reproducibility within and between labs

Dye reversal

Pooled v. control reference mRNA

Control spots from another species

• spiking labeled probes
  • cDNA
  • RNA

• Produced 10 PCR products from yeast genome and 1 from fly genome
• PCR products matched for GC content and length
• Made 10 probe pairs; 2 for each gene (two colors/gene)
• Each probe matched for GC content and length
• Probes calculated not to bind to wrong genes
• Spotted range of concentrations for each gene

 

Unexpected Gene-specific Variation

 

Compared Three Detection Variations

 

---

 

Wanted to Improve Quality of Signal: Clone PCR Products

Original PCR Products	Cloned PCR Products

 

Optimized Experimental Conditions: Signal Up and Noise Down

 

What about Other Ratios? 10:1, 3:1, 1:1, 1:3, 1:10

---

Good Trends

• Reproduce each experiment at least 3 times for each dye (total of 6 times).
  
• Spot oligos if possible to minimize unintentional cDNA-to-spot binding.
  
• Verify a few genes (how many?) by alternative methods (real-time PCR, Northern blots, RNase protection assay, etc.).
  
• Use DNA microarrays to detect candidate genes and overall trends. Do NOT use them to quantify activity of individual genes.