Related Experiment Video
Updated: May 10, 2026
A Data Integration Workflow to Identify Drug Combinations Targeting Synthetic Lethal Interactions
Published on: May 27, 2021
Lit-OTAR framework for extracting biological evidences from literature.
Santosh Tirunagari1, Shyamasree Saha1, Aravind Venkatesan1
1Literature Services Team, European Bioinformatics Institute, European Molecular Biology Laboratory (EMBL-EBI), Wellcome Trust Genome Campus, Cambridge CB10 1SD, United Kingdom.
The lit-OTAR framework uses deep learning to extract evidence from scientific literature, accelerating drug discovery by identifying novel drug targets and validating associations. This system processes millions of articles, revealing millions of target-disease, target-drug, and disease-drug relationships.
Area of Science:
- Biomedical Informatics
- Drug Discovery
- Computational Biology
Background:
- Drug target identification and validation are critical but complex processes.
- Extracting actionable insights from vast scientific literature is challenging.
Purpose of the Study:
- To develop and implement a deep learning framework (lit-OTAR) for automated evidence extraction from scientific literature.
- To enhance drug discovery by identifying and validating potential drug targets and their associated relationships.
Main Methods:
- Utilized deep learning, specifically named entity recognition (NER) and entity normalization.
- NER identifies gene/protein, disease, organism, and chemical/drug entities in scientific texts.
- Entity normalization maps identified entities to established databases (Ensembl, EFO, ChEMBL).
Main Results:
- Processed over 39 million abstracts and 4.5 million full-text articles and preprints.
- Identified over 48.5 million unique associations.
- Discovered >29.9 million target-disease, 11.8 million target-drug, and 8.3 million disease-drug relationships.
Conclusions:
- The lit-OTAR framework significantly accelerates drug discovery and scientific research.
- Automated literature evidence extraction provides a scalable solution for identifying drug-target associations.
- The framework's continuous operation and extensive data processing demonstrate its utility and impact.
