Knowledge graphs (KGs) represent the connections and relationships between real-world entities. We propose a link prediction framework on KGs named Enrichment-Driven GrAph Reasoner (EDGAR) that infers new edges by mining entity-local rules. This approach is based on enrichment analysis, a well-established statistical method used to calculate mechanisms common to a set of differentially expressed genes. EDGAR’s inference results are inherently explainable and rankable, equipped with p-values for statistical significance of each enrichment-based rule. We demonstrate its effectiveness on a large-scale biomedical KG, ROBOKOP, focusing on drug repurposing for Alzheimer disease (AD) as a case study. Initially, we extracted 14 known drugs from the KG and identified 20 contextual biomarkers through enrichment analysis, shedding light on functional pathways relevant to the shared efficacy of drugs for AD. Subsequently, using the top 1,000 enrichment results, our enrichment-driven system identified 1,246 additional drug candidates for AD treatment. We validated the top 10 candidates using medical literature evidence. EDGAR is deployed within ROBOKOP, along with a web user interface. This is the first work to use enrichment analysis for either large graph completion or drug repurposing.

This paper aims to accelerate scientific discovery by studying link prediction in a semantic network. The nodes are unidentified concepts in machine learning, and the time-stamped edges indicate co-occurrence in scientific papers. Taking advantage of this temporal information, we perform node embedding on the graph at every year from 1994 to 2017, and apply two methods to find features for node pairs: the first method uses a transformer, while the other uses distance metrics combined with known link prediction features. The latter feature extraction technique with a 3-layer multi-layer perceptron achieved an AUC of 0.902 on predicting edges in the 2020 graph. Inspection of the resulting features suggests that the model does indeed pay attention to the dynamic nature of the features, e.g., how node-pair distance in embedding space changes over the years.