Development of a classification scheme for disease-related enzyme information
2011

DRENDA: A Database for Disease-Related Enzyme Information

Sample size: 910897 publication Evidence: moderate

Author Information

Author(s): Söhngen Carola, Chang Antje, Schomburg Dietmar

Primary Institution: Technische Universität Braunschweig

Hypothesis

Can we develop a classification scheme for disease-related enzyme information using text mining and machine learning?

Conclusion

DRENDA successfully categorizes enzyme-disease relationships, enhancing the understanding of their interactions.

Supporting Evidence

  • DRENDA categorizes enzyme-disease relationships into four categories: causal interaction, therapeutic application, diagnostic usage, and ongoing research.
  • The classification achieved an F1 score between 0.802 and 0.738 depending on the category.
  • DRENDA is biannually updated to include new enzyme-disease relationships.

Takeaway

DRENDA helps scientists find out how enzymes are related to diseases by organizing information from research papers.

Methodology

The study used text mining and machine learning to classify enzyme-disease relationships from PubMed abstracts.

Potential Biases

Potential biases may arise from the reliance on existing literature and the limitations of text mining techniques.

Limitations

The study may miss some enzyme-disease relationships due to the complexity of language and the need for manual curation.

Digital Object Identifier (DOI)

10.1186/1471-2105-12-329

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication