Semantic rules for extracting proteins functions information from biomedical abstracts

We present a classifier system called SRPFP that predicts the functions of un-annotated proteins. SRPFP aims at enhancing the state of the art of biological text mining. It analyzes biomedical texts in order to discover protein function information that is difficult to retrieve. It employs semantic rules for extracting proteins functions information from biomedical abstracts. It applies a novel model and linguistic computational techniques for extracting the functional relationship from different structural forms of terms in the sentences of biological abstracts. Specifically, SRPFP extracts phrases that represent functional relationships between proteins and molecules. These molecules usually bind to the proteins and are highly predictive of the functions of these proteins. The proposed semantic rules can identify the semantic relationship between each co-occurrence of a protein-molecule pair using the syntactic structures of sentences and linguistics theories. SRPFP represents each protein by the molecules that have high co-occurrences with the protein in biomedical abstracts. This is because such molecules are good characteristics and indicators of the functions of proteins. SRPFP measures the semantic similarity between the molecules representing an un-annotated protein p and the molecules representing annotated proteins and assigns p the functions of annotated proteins that are similar to p.