OpenBioLink: A benchmarking framework for large-scale biomedical link prediction

TL;DR

OpenBioLink introduces a comprehensive benchmark dataset for evaluating large-scale biomedical link prediction algorithms, enabling transparent and reproducible assessment of machine learning methods in biomedical knowledge networks.

Contribution

It provides the first large-scale, high-quality benchmark framework specifically designed for biomedical link prediction tasks, along with baseline evaluation results.

Findings

Benchmark dataset available for research use

Baseline evaluation results demonstrate the framework's utility

Supports development of improved biomedical link prediction algorithms

Abstract

SUMMARY: Recently, novel machine-learning algorithms have shown potential for predicting undiscovered links in biomedical knowledge networks. However, dedicated benchmarks for measuring algorithmic progress have not yet emerged. With OpenBioLink, we introduce a large-scale, high-quality and highly challenging biomedical link prediction benchmark to transparently and reproducibly evaluate such algorithms. Furthermore, we present preliminary baseline evaluation results. AVAILABILITY AND IMPLEMENTATION: Source code, data and supplementary files are openly available at https://github.com/OpenBioLink/OpenBioLink CONTACT: matthias.samwald ((at)) meduniwien.ac.at