Analyze miRNA microarrays

Question

[madkitty]

We sequenced miRNA on Agilent microarrays from different multiple cell lines. Our data are stored on a text file. I've looked over the literature and see a wide variety of inconsistent normalization method (log 2, log 2 and 90th percentile, 95th percentile, no normalization.. )

I'm not familiar with normalization protocols and tools to use to take care of microarrays of miRNA, what tool do you guys use and which type of normalization do you do on miRNA data?

1. TOOLS TO ANALYZE MIRNA (Heatmap, enrichment etc..)
2. NORMALIZATION

Thanks!

Answer 1

[Lorena Pantano]

Hi,

I am not an expert but starting with this package may help: http://master.bioconductor.org/packages/release/bioc/vignettes/AgiMicroRna/inst/doc/AgiMicroRna.pdf

A good normalization will show similar intensity distribution and clustering the samples by replicates for instance. In arrays in general you do noise subtraction (meaning removing noise signal), and then quantile normalization or median normalization, and then log2 transformation. But it depends of what you data needs, and the density, boxplot, heatmap, PCA ... help to decide this.

hope it helps.

[madkitty]

Thanks a lot for your answer! I feel like I need a whole course 101 on how to analyze miRNA microarray...

Thanks!