VitaGraph: Building a Knowledge Graph for Biologically Relevant Learning Tasks

TL;DR

VitaGraph addresses the challenge of creating a reliable, richly featured knowledge graph for biologically relevant learning tasks by starting from the Drug Repurposing Knowledge Graph (DRKG) and applying a rigorous cleaning pipeline, integrating Reactome pathways and OnSIDES drug-side effect data, and enriching nodes with Morgan fingerprints and gene-function features. The authors benchmark VitaGraph on three core link-prediction tasks—drug repurposing, PPI prediction, and side-effect identification—demonstrating competitive performance while revealing data leakage issues in the original DRKG. They present a configurable pipeline that extends DRKG with additional data sources and meaningful features, providing a reproducible resource and benchmark for graph neural networks in network medicine. The work highlights VitaGraph’s potential to accelerate biologically informed link prediction and precision medicine research, while acknowledging dependence on source data quality and the need for regular updates and validation of new connections.

Abstract

The intrinsic complexity of human biology presents ongoing challenges to scientific understanding. Researchers collaborate across disciplines to expand our knowledge of the biological interactions that define human life. AI methodologies have emerged as powerful tools across scientific domains, particularly in computational biology, where graph data structures effectively model biological entities such as protein-protein interaction (PPI) networks and gene functional networks. Those networks are used as datasets for paramount network medicine tasks, such as gene-disease association prediction, drug repurposing, and polypharmacy side effect studies. Reliable predictions from machine learning models require high-quality foundational data. In this work, we present a comprehensive multi-purpose biological knowledge graph constructed by integrating and refining multiple publicly available datasets. Building upon the Drug Repurposing Knowledge Graph (DRKG), we define a pipeline tasked with a) cleaning inconsistencies and redundancies present in DRKG, b) coalescing information from the main available public data sources, and c) enriching the graph nodes with expressive feature vectors such as molecular fingerprints and gene ontologies. Biologically and chemically relevant features improve the capacity of machine learning models to generate accurate and well-structured embedding spaces. The resulting resource represents a coherent and reliable biological knowledge graph that serves as a state-of-the-art platform to advance research in computational biology and precision medicine. Moreover, it offers the opportunity to benchmark graph-based machine learning and network medicine models on relevant tasks. We demonstrate the effectiveness of the proposed dataset by benchmarking it against the task of drug repurposing, PPI prediction, and side-effect prediction, modeled as link prediction problems.

Figures (3)

• Figure 1: Construction of the gene feature vector. A one-valued entry in the feature vector means that the gene was connected to a given entity of that type in the original DRKG.
• Figure 2: Comparison of the original DRKG and the new VitaGraph schema. In VitaGraph schema only a limited number of node and edge types are preserved, augmenting the coherence and reducing the noise within the graph.
• Figure 3: VitaGraph creation pipeline.