Biomedical Information Extraction for Disease Gene Prioritization

TL;DR

This paper presents a biomedical information extraction pipeline that enhances disease-gene association predictions by extracting and integrating biological relationships from vast PubMed texts, outperforming existing methods.

Contribution

It introduces a novel IE pipeline for extracting biological relationships and demonstrates its effectiveness in augmenting knowledge graphs for disease-gene prediction.

Findings

Improved disease-gene prediction accuracy by 20% hit@30.

Enhanced knowledge graph with new protein-protein interactions.

Outperformed state-of-the-art BioNLP components.

Abstract

We introduce a biomedical information extraction (IE) pipeline that extracts biological relationships from text and demonstrate that its components, such as named entity recognition (NER) and relation extraction (RE), outperform state-of-the-art in BioNLP. We apply it to tens of millions of PubMed abstracts to extract protein-protein interactions (PPIs) and augment these extractions to a biomedical knowledge graph that already contains PPIs extracted from STRING, the leading structured PPI database. We show that, despite already containing PPIs from an established structured source, augmenting our own IE-based extractions to the graph allows us to predict novel disease-gene associations with a 20% relative increase in hit@30, an important step towards developing drug targets for uncured diseases.