Affymetrix is a microarray manufacturer: http://www.affymetrix.com

Affymetrix manufactures one color microarrays. Their chips consist of small DNA fragments, referred to as probes, synthesized at specific locations (spots) on a coated quartz surface. Every probe in an Affymetrix microarray has 25 nucleotide length. It is designed to determine whether or not the complementary sequence of an RNA or DNA molecule is present in the sample, and to quantify its concentration.

Affymetrix uses a set probes to interrogate each genomic feature like genes or exons. Such group of probes is called probe-set and can be averaged into a single measurement for the genomic feature.

In classical Affymetrix arrays, for each probe that perfectly matches (PM) its target sequence, Affymetrix also builds a paired mismatch probe (MM). The mismatch probe contains a change in the nucleotide located directly in the middle of the 25-base probe sequence and is designed to measure non specific hybridization. A prefect match probe and its mismatch are referred to as a probe-pair. Thanks to the MM probes, Affymetrix microarrays can be used to get a present-absent call for each gene, indicating whether it is present or not in the sample.

Recently, Affymetrix has refined its technology and, among other changes, new microarrays do not follow the PM-MM design. In contrast, a group of probes which do not match biological sequences is used as control for all other PM probes. This design increases the amount of features that can be screened while keeping the possibility of getting present-absent calls for them.

After scanning each array, Affymetrix software computes probe intensity data and saves them into a CEL file using a particular structure. Such CEL files are the starting point for the Affymetrix normalization module in Babelomics: your raw data.

If you are using an standard Affymetrix chip our tool will recognize the kind of array you are using and you will be able to normalize it. At the moment it is not possible to analyze custom Affymetrix arrays using Babelomics.

Many different algorithms have been proposed in the past years for normalizing Affymetrix microarray data. At the end, two of them become standards and are the ones available in Babelomics:

• RMA
• PLIER