R interface to the reactome graph database

Chi Lam Poon, Justin Cook, Solomon I. Shorser et al. · 2021

Reactome is a free, open-source, curated, and peer-reviewed knowledgebase of biomolecular pathways. As of the current release, its holdings comprise 13,827 human reactions organized into 2,536 pathways involving 11,084 proteins, 1,857 small molecules, and 432 drugs, with the support of 33,752 literature references. These molecular entity and event annotations are systematically represented as an ordered network of molecular reactions, forming a graph of biological knowledge. Reactome has migrated from a relational database to a Neo4j graph database for efficient storage, processing, and access with these highly interconnected data. To minimize the usage of Neo4j Cypher queries, we developed an interface, ReactomeGraph4R, for interacting between R and a local Reactome graph database. The ReactomeGraph4R package retrieves data with a network structure and contains functions that permit complex tasks, such as i) finding hierarchical data of an instance; ii) getting the entire Reaction context using preceding or following relationships; and iii) displaying network graphs. Output data are formatted as R objects to assist with downstream analyses. We developed a second R package, ReactomeContentService4R, to wrap the RESTful API of the Content Service built on top of the graph core. These packages are available through Bioconductor and fulfill the needs of Reactome pathway- and network-based tools for analyzing multiple omics datasets.

Cited by 4 works. doi:10.7490/f1000research.1118690.1

Authors