A Novel Approach for Protein-Named Entity Recognition and Protein-Protein Interaction Extraction

Figure 2

User interface of the proposed ProNER model and PPI extraction model. (a) Interface of ProNER for biomedical sentence input. Here, users can select the Protein-NER methods, dictionary-based method, or data mining-based method. Users execute the PPIMiner by using the button “Identify.” (b) Final result of Protein-NER. Here, the “named entity-mapping” represents the recommendation name of this protein or gene name or other named entities that were extracted using data mining-based (machine learning-based) method. (c) Interface of PPI extraction system for inputting biomedical sentence. User can also select the Protein-NER method. User can search the protein-protein interactions from our PPI database. (d) PPI patterns searched by proposed model. From the result page, we can get the information of PPIs, the type of PPIs, PubMed ID of resource documents, and the full abstract of the resource documents.

(a)

(b)

(c)

(d)