BioKG

Abstract

Knowledge graphs became a popular means for modelling complex biological systems where they model the interactions between biological entities and their effects on the biological system. They also provide support for relational learning models which are known to provide highly scalable and accurate predictions of associations between biological entities. Despite the success of the combination of biological knowledge graph and relation learning models in biological predictive tasks, there is a lack of unified biological knowledge graph resources. This forced all current efforts and studies for applying a relational learning model on biological data to compile and build biological knowledge graphs from open biological databases. This process is often performed inconsistently across such efforts, especially in terms of choosing the original resources, aligning identifiers of the different databases and assessing the quality of included data. To make relational learning on biomedical data more standardised and reproducible, we propose a new biological knowledge graph which provides a compilation of curated relational data from open biological databases in a unified format with common, interlinked identifiers. We also provide a new module for mapping identifiers and labels from different databases which can be used to align our knowledge graph with biological data from other heterogeneous sources. Finally, to illustrate practical relevance of our work, we provide a set of benchmarks based on the presented data that can be used to train and assess the relational learning models in various tasks related to pathway and drug discovery.

Ensembles of knowledge graph embedding models improve predictions for drug discovery

“…We benchmarked two KGs: BioKG [ 32 ] and OpenBioLink [ 33 ]. Since both KGs are designed for a variety of biomedical applications (e.g., drug repurposing and side effect predictions), they contain different node (e.g., proteins, phenotypes and anatomical regions) and relation types (e.g., inhibition, activation and binding in OpenBioLink, and protein–protein interactions and drug–drug interactions in BioKG) that were normalized in the steps outlined below.…”

A knowledge graph representation learning approach to predict novel kinase–substrate interactions

“…We construct the knowledge graph by including data from iPTMnet, 3 Protein Ontology (PRO), 4 Gene Ontology (GO) 5 and BioKG. 6 To begin with, we use human PTM data [Taxon code – 9606] from iPTMnet to construct a kinase–substrate interaction network. The iPTMnet data contains 26411 phosphorylation PTM events.…”

“…Rather than performing this integration ourselves, we decided to take advantage of the BioKG database. 6 The authors of BioKG database provide a framework to automatically integrate data from numerous biomedical databases. Since BioKG framework is geared towards drug discovery analysis we integrated only a subset of the biomedical databases.…”

“…To this date, many tools have been developed to predict kinase–substrate interactions. Tools such as Scansite, 6 NetPhospK, 7 PPSP, 8 GPS 9,10 and PredPhosph 11 rely on the properties of protein sequences around the phosphorylation site also known as “sequence motifs”, to predict kinases most likely to be associated with the given phosphorylation site. But kinase–substrate interactions involve much more than sequence motifs and hence it is necessary to include contextual factors when making these predictions.…”

Utilizing graph machine learning within drug discovery and development

“…Biomedical databases are typically presented as general repositories with minimal reference to downstream applications. Therefore, additional processing is often required for a specific task; to this end, efforts have been directed towards processed data repositories with specific endpoints in mind [ 115 , 116 ].…”