BioRED: A Rich Biomedical Relation Extraction Dataset

TL;DR

BioRED is a comprehensive biomedical relation extraction dataset with multiple entity types and relation pairs at the document level, enabling improved development and benchmarking of RE systems in biomedicine.

Contribution

It introduces BioRED, the first biomedical RE corpus with diverse entity types, relation pairs, and relation annotations for novel and background knowledge, at the document level.

Findings

High NER performance (F1=89.3%) achieved by existing models.

RE performance remains moderate (F1=47.7%) especially for novel relations.

Rich dataset facilitates development of more accurate biomedical RE systems.

Abstract

Automated relation extraction (RE) from biomedical literature is critical for many downstream text mining applications in both research and real-world settings. However, most existing benchmarking datasets for bio-medical RE only focus on relations of a single type (e.g., protein-protein interactions) at the sentence level, greatly limiting the development of RE systems in biomedicine. In this work, we first review commonly used named entity recognition (NER) and RE datasets. Then we present BioRED, a first-of-its-kind biomedical RE corpus with multiple entity types (e.g., gene/protein, disease, chemical) and relation pairs (e.g., gene-disease; chemical-chemical) at the document level, on a set of 600 PubMed abstracts. Further, we label each relation as describing either a novel finding or previously known background knowledge, enabling automated algorithms to differentiate between novel and background information. We assess the utility of BioRED by benchmarking several existing state-of-the-art methods, including BERT-based models, on the NER and RE tasks. Our results show that while existing approaches can reach high performance on the NER task (F-score of 89.3%), there is much room for improvement for the RE task, especially when extracting novel relations (F-score of 47.7%). Our experiments also demonstrate that such a rich dataset can successfully facilitate the development of more accurate, efficient, and robust RE systems for biomedicine. The BioRED dataset and annotation guideline are freely available at https://ftp.ncbi.nlm.nih.gov/pub/lu/BioRED/.