Differences between chip type and chip definition file (CDF)

Chip type

Each Affymetrix microarray has a particular chip type.  The chip type identifies the type array produced in the factory.  The chip type and its content, such as the probes, the probe sequences, the probe locations, will never change during the life time of the array and array type.  

Example of chip types are Mapping10K_Xba142, GenomeWideSNP_6, HG-U133A, and HuGene-1_0-st-v1.  

In early stages, Affymetrix sometimes labels their chip types different.  For instance, early on the chip type GenomeWideSNP_6 was also referred to as GenomeWideEx_6 and GenomeWideEx_6_v2.

 

Chip definition file (CDF)

A CDF is an annotation file that contains one (of many possible) annotations of a particular chip type.  A CDF typically specifies which probes (cells) maps to the same genomic unit of interest.  For instance, for gene expression there are CDFs that specifies sets of probes that maps to the same gene, and for genotyping arrays there are CDFs that specifies which probes interrogates which SNP and which allele type.  Different CDFs may be used to interrogate different genomic units.  For instance, one CDF may be used to interrogate gene transcripts and another CDF to interrogate exons, although the two CDFs are for the same chip type.

Affymetrix provides "official" CDFs for most of their chip types.  In addition to these, various groups provides so called "custom CDFs" that are optimized for various genomic features.  CDFs are typically updated when the genome annotation is updated.

Example of CDFs are Mapping10K_Xba142.cdf, GenomeWideSNP_6.cdf, GenomeWideSNP_6.Full.cdf, HuEx-1_0-st-v2.cdf and HuEx-1_0-st-v2,coreR3,A20071112,EP.cdf.

 

Notes

Note 1: Note that the term 'CDF' is sometimes used incorrectly when the term 'chip type' is meant.  

Note 2: In the Bioconductor project, there are 'CDF environments' and 'CDF packages'.  Although sometimes referred to as "CDFs" in the Bioconductor community, these are not CDFs, but instead R-specific data structures that contain all or parts of the information available in a CDF.