Understanding the Performance of Knowledge Graph Embeddings in Drug Discovery

TL;DR

This paper investigates how various factors like training setup, hyperparameters, and data splits influence the performance of knowledge graph embedding models in drug discovery, emphasizing the need for comprehensive reporting for reproducibility.

Contribution

It provides a detailed analysis of the impact of experimental factors on KGE model performance in drug discovery, highlighting the importance of reporting these factors for reproducibility.

Findings

Training setup significantly affects model performance.

Hyperparameters and initialisation seed influence results.

Performance rankings of models can change based on experimental choices.

Abstract

Knowledge Graphs (KG) and associated Knowledge Graph Embedding (KGE) models have recently begun to be explored in the context of drug discovery and have the potential to assist in key challenges such as target identification. In the drug discovery domain, KGs can be employed as part of a process which can result in lab-based experiments being performed, or impact on other decisions, incurring significant time and financial costs and most importantly, ultimately influencing patient healthcare. For KGE models to have impact in this domain, a better understanding of not only of performance, but also the various factors which determine it, is required. In this study we investigate, over the course of many thousands of experiments, the predictive performance of five KGE models on two public drug discovery-oriented KGs. Our goal is not to focus on the best overall model or configuration, instead we take a deeper look at how performance can be affected by changes in the training setup, choice of hyperparameters, model parameter initialisation seed and different splits of the datasets. Our results highlight that these factors have significant impact on performance and can even affect the ranking of models. Indeed these factors should be reported along with model architectures to ensure complete reproducibility and fair comparisons of future work, and we argue this is critical for the acceptance of use, and impact of KGEs in a biomedical setting.