Among the articles in this special issue is a review on the current status of techniques for mining of the pharmacogenomics literature and includes descriptions of studies that have utilized the corpus of human-curated pharmacogenetic (PGx) articles in PharmGKB to train or assess the accuracy of machine-learning methods.
"The PharmGKB repository comes perhaps closest to the vision of an all-embracing pharmacogenomics corpus. It represents a major step towards an interdisciplinary biomedical information store." quote from:
Mining the pharmacogenomics literature -- a survey of the state of the art. Hahn U, Cohen KB, Garten Y, Shah NH. Brief Bioinform. 2012 Jul;13(4):460-94.
At the PharmGKB we are currently working on a natural language processing (NLP) pipeline that would assist in the identification of relevant PGx articles for our curators to then manually curate.