scribl: A system for the semantic capture of relationships in biological literature

In systems biology research, utilizing literature databases involves more than simple keyword queries for biological agents (e.g., proteins, genes, compounds, receptor complexes) and processes (e.g, autophagy, cell cycle) (Krallinger et al., 2008), which typically only return lists of articles.Advanced methods are necessary for extracting and visualizing the relationships detailed within these documents (Cary et al., 2005;Pavlopoulos et al., 2015;Suderman & Hallett, 2007).Here, we introduce a system that supports the annotation of scientific articles and represents and visualizes these relationships.This system, scribl, consists of two parts:(1) a simple syntax that can be used to curate the biological relationships described within the text of those articles, (2) a Python software API and pipeline that can transform a Zotero literature database, with entries annotated with this syntax, into a database suitable for graph-based relationship queries.