Identification of protein interaction methods from biomedical literature

Proteins are the functional subunits of a cell which interact with each other to carry out biological processes. Protein interaction networks form the backbone of the research in molecular and systems biology. Although there are available methods to mine protein interactions and their detection methods from the biological literature, the accuracy of these methods is quite low. In this study, we applied regular expressions to identify the three most frequent protein interaction detection methods from the methodology section of the full text articles. These articles were then further used to extract the protein protein interactions. We report an overall specificity of 83.6 and sensitivity of 78.2 for the identification of interaction methods.