Identification of differentially expressed proteins using automated meta-analysis of proteomic articles

An automated method for searching for differentially expressed proteins (DEP) in proteomic articles has been developed and tested. Full-text proteomics-related articles were selected using an electronic version of the journal Proteomics and PubMedCentral. The list of proteins most frequently mentioned in the articles shares 86% identity with the list of human frequently identified DEPs published recently (Petrak et al., Proteomics (2008) 8, 1744). Regardless of the goal or design or methods identification of DEP delivers annexins and peroxiredoxins, as well as alpha-enolase, triosephosphate isomerase, and heat-shock protein HSP60. Among the most often mentioned proteins were also serum albumin, cathepsin D and vimentin. In regard to protein function the most often mentioned proteins were involved in inflammation and the immune response. According to GenRIF and UniProtKB annotations, most of these proteins are linked to the pathogenesis of tumor diseases, or diseases of the cardiovascular and nervous systems.