Dexter Pratt is the director of the NDEx (Network Data Exchange) project at UCSD, part of the Ideker lab and closely associated with the Cytoscape project. NDEx (www.ndexbio.org) is an open source software framework and a public website for sharing, publication, and application use of biological networks. Dexter was the original author of BEL (Biological Expression Language) in 2003 at Selventa, Inc. (formerly Genstruct). He was also an architect in the redesign of BEL as it evolved from a proprietary technology to the open standard OpenBEL (www.openbel.org). Dexter has more than 30 years of commercial experience in knowledge representation, semantic databases and systems biology databases. His earlier work includes software and hardware development on the pioneering Lisp Machine AI workstations followed by research at the Cyc commonsense reasoning project.
Title: PyBEL: A BEL graph analysis software package
OpenBEL is an attractive language for the expression of scientific findings in biology and has grown in use since its public release. It is concise, focused on causal relationships, and, for many biologists, it is intuitive. It promotes the capture of durable facts and their assembly into task-specific models. In this perspective on next steps for BEL, I will present three areas which I think should be priorities for the BEL community. First, we need interoperability, both in collaboration and convergence with other causal representation efforts and in creating BEL-derived models in formats easily consumed by the systems biology and bioinformatics communities. Second, we should promote the knowledge assembly paradigm by the creation and dissemination of easily used tools and services. Finally, we should leverage key strengths of BEL in expressing multi-scale experimental observations. The expression of assertions that are based directly in perturbations and measurements yields the most durable and reusable facts. BEL provides a means to capture observations from diverse experiments in a common form, providing a firm footing for integrated, systematic analysis of multi-omic data. In sum, BEL and its infrastructure should evolve to meet the needs of automated biological reasoning across a much larger community, becoming an essential part of “BD2K” – Big Data to Knowledge