• Exon skipping
• Intron retention
• Mutually exclusive exon usage
• Alternative promoter usage
• Alternative polyadenylation
• Alternative splicing donor/acceptor site with changes over 25 bp

Combining empirical sequences and gene predictions for novel discoveries

The Exon Array supports the most detailed survey of the genome, and expands the discovery of novel transcript structures beyond the well-characterized and annotated genes. Affymetrix took an all-inclusive design approach and the Exon Arrays contain sequences from two primary sources:

• cDNA-based content including the more established RefSeq mRNAs, GenBank® mRNAs, and ESTs from dbEST. In addition, mouse and rat cDNAs have been mapped onto the human genome using genome synteny maps from the UCSC Genome Bioinformatics group.



• Predicted gene structure sequences from:
  
  
• GENSCAN: genes.mit.edu/GENSCAN.html
• Ensembl: www.ensembl.org/
• Vega: vega.sanger.ac.uk/
• geneid and sgp: www1.imim.es/software/geneid/index.html
• TWINSCAN: genes.cs.wustl.edu/
• Exoniphy: genome.ucsc.edu/cgi-bin/hgTrackUi?g=exoniphy&db=hg16
• MicroRNA Registry: www.sanger.ac.uk/Software/Rfam/mirna/
• MITOMAP: www.mitomap.org/
• Structural RNA Predictions: genome.ucsc.edu/cgi-bin/hgTrackUi?g=rnaGene&db=hg16
  
  

One probe set for each exon for complete coverage

Most annotated exons or exon elements that are over 25 bp in length are represented with at least one probe set.

Figure 1. Schematic for coverage of probe sets across the entire length of the transcript. Golden regions are exons whereas the grey regions represent introns that are removed during splicing. The short dashes underneath the exon regions for the Exon Array and 3'-Array PSR (probe selection region) indicate individual probes representing that PSR.

Anchoring the design on the genome to correlate expression data with genetic information

Each probe on the array is designed from the genomic sequence and is annotated with its genome coordinates. Projecting the design onto the genome facilitates the dynamic update of the design reflecting the new knowledge of the genomic sequence and annotations.

In addition, this design and annotation approach also enables more convenient correlation of the expression data with DNA sequence information.

Syntenic design for a path to comparative genomics

In addition to the human cDNAs and ESTs, the full-length mouse and rat cDNAs were mapped to their respective genomes and the syntenic information was included in the design of the Human Exon Array.

Supplementing the design with mouse and rat sequences allows more direct examination of splicing events across the three key organisms, providing information on the evolution and conservation of exon utilization and expression.

One million exons on a single array

Two key technological advancements have enabled this exon array design, providing cost and sample savings for analyzing the whole genome on only one array:

• Highest density array manufacturing at 5µm feature size
• More efficient background subtraction method utilizing surrogate mismatch probes