Schematic of mapping strategies. (1) Conventional approach. A summary value for each protein is mapped to its corresponding plus2 array probeset via SWISS-PROT sequence identifiers. (2–4) Mappings performed via the genome. (A) Exon array and plus2 array probesets are associated with their target exons using the X:Map database. These are then mapped to genomic coordinates via X:Map and Ensembl. (B) Peptide sequences are mapped to proteins using a sequence search against the Ensembl protein database and then mapped to genomic coordinates using the Ensembl Perl API. (2) Exon array comparison. Individual peptides are mapped to exons and then to their corresponding exon array probeset. (3) Plus2 array comparison. Individual peptides are mapped to exons and then to array probesets, as before. Since the majority of transcripts on a plus2 array are represented only by a single probeset placed at the 3' end of the transcript, many peptides cannot be mapped to an appropriate probeset. (4) Exon array mappings. A summary fold-change is created for the proteomics and microarray data by averaging across the peptides or probesets, respectively. These are then mapped via the genome to produce a comparison between transcript and protein level summaries.
Bitton et al. BMC Bioinformatics 2008 9:118 doi:10.1186/1471-2105-9-118