Should I choose RNA-seq or Microarray for my future experiment?

[creativediagnostics]

If I have for example 5 to 10 samples of a few different developmental time periods that I want to compare gene expression, what should be my pick for accuracy? Splicing info is interesting but not a deal breaker.

[creativediagnostics]

Thanks in advance!

[chapmandu2]

Depends on your experiment and your situation! Do you have access to RNA-seq and/or microarray in house or will you outsource? Is there a bioinformatics group who can at least do the sequence alignment and read counting part of the analysis for you? Personally I can't see why I'd ever do microarray again because I can get RNA-seq done in house cheaply and easily, there's a pipeline to do the alignment, and I've analysed it before. That's before you consider the benefits of dynamic range, sensitivity etc. OTOH, the benefit of microarray is that you get nice easy CEL files back and the analysis is easy.

Take care of your experimental design... Replicates are important to estimate within-group variance. Don't bother doing paired end seq if you just want to measure gene level expression, single end is cheaper per read and so more cost effective. No need to go above 15-20M reads per sample unless you're really interested in low expressers, spend the money on more reps instead.