Related Experiment Video
Updated: May 13, 2026
14:53
In Situ Detection and Single Cell Quantification of Metal Oxide Nanoparticles Using Nuclear Microprobe Analysis
Published on: February 3, 2018
Using nanoinformatics methods for automatically identifying relevant nanotoxicology entities from the literature
Miguel García-Remesal1, Alejandro García-Ruiz, David Pérez-Rey
1Departamento de Inteligencia Artificial, Facultad de Informática, Universidad Politécnica de Madrid, Boadilla del Monte, 28660 Madrid, Spain. mgarcia@infomed.dia.fi.upm.es
Biomed Research International
|March 20, 2013
Summary
Nanoinformatics aids nanotoxicology by automatically extracting key data like nanoparticles, exposure routes, and toxic effects from scientific literature. This computational approach accelerates nanomedicine research and knowledge management.
Area of Science:
- Nanoinformatics
- Computational Toxicology
- Biomedical Informatics
Background:
- Nanoinformatics leverages informatics for data management in nanotechnology, particularly for healthcare applications.
- Nanotoxicology research faces challenges in efficiently processing vast amounts of scientific literature.
- Automated data extraction is crucial for advancing nanomedicine and understanding nanoparticle safety.
Purpose of the Study:
- To develop and evaluate a computational approach for extracting nanotoxicology-related entities from scientific literature.
- To demonstrate the feasibility of using nanoinformatics to facilitate nanotoxicology research.
- To identify key entities including nanoparticles, routes of exposure, toxic effects, and targets.
Main Methods:
- Developed a named entity recognition (NER) system using a custom-created corpus.
- Trained the NER system on four categories: nanoparticles, routes of exposure, toxic effects, and targets.
- Validated the system with nanomedicine/nanotoxicology experts and evaluated using 10-fold cross-validation.
Main Results:
- Achieved high precision ranging from 87.6% (targets) to 93.0% (routes of exposure).
- Recalled values ranged from 82.6% (routes of exposure) to 87.4% (toxic effects).
- Demonstrated the reliability of computational methods for NER-dependent tasks in nanotoxicology.
Conclusions:
- Computational approaches, specifically NER, are feasible for reliably extracting critical nanotoxicology information.
- This proof-of-concept can be expanded to enhance data management and accelerate nanomedicine research.
- Automated extraction supports tasks like augmented reading and semantic searches, improving knowledge discovery.
