Improvements in the protein identifier cross-reference service

The Protein Identifier Cross-Reference (PICR) service is a tool that allows users to map protein identifiers, protein sequences and gene identifiers across over 100 different source databases. PICR takes input through an interactive website as well as Representational State Transfer (REST) and Simple Object Access Protocol (SOAP) services. It returns the results as HTML pages, XLS and CSV files. It has been in production since 2007 and has been recently enhanced to add new functionality and increase the number of databases it covers. Protein subsequences can be Basic Local Alignment Search Tool (BLAST) against the UniProt Knowledgebase (UniProtKB) to provide an entry point to the standard PICR mapping algorithm. In addition, gene identifiers from UniProtKB and Ensembl can now be submitted as input or mapped to as output from PICR. We have also implemented a 'best-guess' mapping algorithm for UniProt. In this article, we describe the usefulness of PICR, how these changes have been implemented, and the corresponding additions to the web services. Finally, we explain that the number of source databases covered by PICR has increased from the initial 73 to the current 102. New resources include several new species-specific Ensembl databases as well as the Ensembl Genome ones. PICR can be accessed at http://www.ebi.ac.uk/Tools/picr/.