Module B: Extract protein-interaction (PI) information from raw text

Using your/our labeled training data (with labeled PI tripets: two proteins + interaction word) and/or raw query sentences/abstracts. You can train a Bayesian model by using your labeled training data file (leave the field blank if you want to use our default training file). Please refer FAQs for detailed instructions and file formats to be used.

Labeled Training Data:    Default file: trainingdata.txt
Query Data:
  Sentences   Default file: querysentences.txt
  Pubmed Abstracts   Default file: queryabstracts.txt
    Prior class distribution parameter (% true PIs in your query data)?
  Advanced Options:
      output with specific types of interactions
      output interactions with associated Gene Ontology information in sentence List of GO Terms
      output interactions with associated disease information in sentence List of Diseases