Tools, Services and Software

• Whatizit - This Web interface allows access to information extraction modules that have been developed in the Rebholz group and that are available from servers at the EBI. The text input box of the interface accepts any ASCII text. After selection of a module the text is processed and identified terms or relations are marked up in the output text. The interface serves as a demonstrator to some of the tools. If scripted access to the modules is required then please have a look at the Web page explaining the Whatizit pipe.
  
  Please cite: Rebholz-Schuhmann et al., (2007) Text processing through Web services: Calling Whatizit. Bioinformatics, doi:10.1093/bioinformatics/btm557
  
  
  
• EBIMed – This combined information retrieval and information extraction engine processes Medline abstracts in realtime. The user provides one or several keywords as query input. The keywords are used to retrieve Medline abstracts, which are all processed thereafter. The analysis identifies UniProt/Swiss-Prot proteins, GO terms, drugs and species from sentences in the Medline documents. All terms and all sentences with such terms are extracted, ranked, sorted and organised into different lists, which allow for fast access to the sentences and the underlying abstracts. EBIMed is based on the modules implemented for Whatizit.
  
  Please cite: Rebholz-Schuhmann,D., et al. (2007) EBIMed – Text crunching to gather facts for proteins from Medline. Bioinformatics, 23(2):e237-44
  
  
  
• Protein Corral – This information retrieval and information extraction engine processes Medline abstracts to extract protein-protein interactions. The user provides one or several keywords as query input to retrieve Medline abstracts. All abstracts are processed to extract protein-protein interactions from the text. All pairs of proteins are ranked, sorted and organised into different lists, which allow for fast access to the sentences and the underlying abstracts. The information extraction modules of Protein Corral are also available for scripted access through the Whatizit pipe.
  
  
• MedEvi – This search engine has the purpose to select all sentences from Medline that can be aligned with the positions of terms in the intial multi-term query (Permuted concordancer). Users of MedEvi have found the tool useful, for example, to find evidence from the literature that prove whether candidate chemicals are involved in a given metabolic pathway, to identify which proteins regulate a given set of proteins, and to find whether a given multi-term ontology concept actually appears in the literature even with a high degree of syntactic variations.
  
  MedEvi is based on the observation that terms with semantic relations are not found too far from each other. MedEvi further identifies additional keywords of biological and statistical significance from local context of matching occurrences in order to help users reformulate their queries for better results.
  
  
• AcroTagger - Identifies and tags the abbreviations in text with xml tags. If the long-form is given in the text or can be guessed from the document context, then the tag surrounding the abbreviation will contain the expansion's normalised form. The system is written in Java and uses SVM light.