Optimizing Syntax-Patterns for Discovering Protein-Protein-Interactions

Supplementary information

Data sources

• List of nouns indicating interactions (INouns)
• List of verbs indicating interactions (IVerbs)
• Evaluation corpus from BioCreAtIvE, 1000 sentences: please see [here >>]
• Evaluation corpus from DIP references, 297 interactions in 165 papers: please see [here >>]

Software

• POS-Tagging: TnT, Thorsten Brandts, see http://www.coli.uni-sb.de/~thorsten/tnt/
• Protein name recognition: own software, see reference [5] in paper
• Support Vector Machine for [5]: SVMlight, Thorsten Joachims, see http://svmlight.joachims.org/
• WBI's Java Packages for Text Mining: please see [here >>]
  (package: de.hu.wbi.textmining.tasks.ppi)