Knowledge Graphs for drug repurposing: a review of databases and methods

Abstract

Drug repurposing has emerged as a effective and efficient strategy to identify new treatments for a variety of diseases. One of the most effective approaches for discovering potential new drug candidates involves the utilization of Knowledge Graphs (KGs). This review comprehensively explores some of the most prominent KGs, detailing their structure, data sources, and how they facilitate the repurposing of drugs. In addition to KGs, this paper delves into various artificial intelligence techniques that enhance the process of drug repurposing. These methods not only accelerate the identification of viable drug candidates but also improve the precision of predictions by leveraging complex datasets and advanced algorithms. Furthermore, the importance of explainability in drug repurposing is emphasized. Explainability methods are crucial as they provide insights into the reasoning behind AI-generated predictions, thereby increasing the trustworthiness and transparency of the repurposing process. We will discuss several techniques that can be employed to validate these predictions, ensuring that they are both reliable and understandable.

Keywords: artificial intelligence; drug repurposing; explainability; graph networks; knowledge graphs.

Figure 1

Two examples of existing KGs. Left: Bioteque KG [22] where 12 types of node and 67 types of edges are included. A total of 450 thousand biological entities and 30 million relationships are presented in the KG. Right: PharMeBINet KG [23] where 2 869 407 different nodes with 66 labels and 15,883,653 relationships with 208 edge types are included in this KG.

Figure 2

Example of an explanation produced as a subgraph. In this example, our AI model has predicted that Drug A could potentially be used to treat Disease B. A possible subgraph explanation could be that Drug A targets Gene X, which is one the genes that causes Disease B.

Figure 3

Overview of node types and data sources incorporated in various KGs as examined in this review. At the center of the illustration, the KGs are prominently displayed, illustrating their central role and connectivity within the data ecosystem. To the left, the existing data sources included in the studied KGs are listed, highlighting the diverse origins of the information that feeds into these graphs. On the right, the various node types included in the KGs are enumerated, demonstrating the range of entities and relationships modeled within these structures. This comprehensive depiction serves to underline the complexity and the multi-dimensional nature of the KGs, showcasing how they integrate disparate data types to foster enhanced analytical capabilities.