Extraction of Fine-grained Semantic Relations for the Human Variome

Abstract

Genetic information can provide important insights into the detailed disease status of a patient, and guide the selection of the best choice of treatment for an individual. The Variome corpus, a small collection of published articles about inherited colorectal cancer, includes annotations of 11 entity types and 13 relation types capturing the relationship between genetic variation and disease. Due to the richness of these annotations, the corpus provides a good testbed for evaluation of biomedical literature information extraction systems. To date, the corpus has been employed to study the performance of mutation extraction systems but has to our knowledge not previously been tested for deve..