What is a standard approach to binarize microarray gene expression data?

Question

I am using gene expression data from a microarray dataset for the purpose of learning the network structure of genes in the dataset.

Now to do this I need to binarize my data, i.e convert the data for each gene to 0s and 1s. I am new to bioinformatics and am not sure what approach is considered standard for microarray data.

I appreciate your inputs regarding this and I am implementing this in R. So if you know any package that does this please let me know.

Answer 1

Based on the info you provide, ArrayBin R package provides you the necessary tools:

• binarize.array() from ArrayBin, allowing:

Implementation of an adaptive method for binarizing gene expression data on a per-probe basis and demonstrate the superior effectiveness of our method when compared with other, commonly used approaches. This adaptive binarization method can be applied to DNA methylation microarray data, which has implications for cross-platform integration, and can reduce batch effects in the data.

In complement, I'd advise you go through some literature and choose the approach which fits you best. Some suggestions:

• Computational analysis of microarray data
• Binary matrix factorization for analyzing gene expression data
• and probably most relevant for you given your application: Multiscale Binarization of Gene Expression Data for Reconstructing Boolean Networks

Maybe you also want to look into more general concepts of transformation prior to microarray data analysis:

• Microarray data normalization and transformation
• A Tutorial Review of Microarray Data Analysis