Detection and classification of long terminal repeat sequences in plant LTR-retrotransposons and their analysis using explainable machine learning
2024

Detecting and Classifying Plant LTR Retrotransposons Using Machine Learning

Sample size: 176917 publication 10 minutes Evidence: moderate

Author Information

Author(s): Horvath Jakub, Jedlicka Pavel, Kratka Marie, Kubat Zdenek, Kejnovsky Eduard, Lexa Matej

Primary Institution: Faculty of Informatics, Masaryk University

Hypothesis

Can machine learning methods effectively classify and analyze long terminal repeat sequences in plant LTR retrotransposons?

Conclusion

The study demonstrates that machine learning models can successfully identify biologically relevant motifs in LTR sequences, enhancing our understanding of their regulatory roles.

Supporting Evidence

  • The best classification achieved an F1 score of 0.89 for superfamily classification.
  • Machine learning models identified key motifs such as TATA-box sequences in LTRs.
  • Explainability analysis revealed the importance of specific sequence positions in LTR identity.

Takeaway

Scientists used computers to help find and understand important parts of plant DNA that help control how genes work.

Methodology

The study employed machine learning techniques, including Gradient Boosting, CNN-LSTM, and DNABERT models, to classify LTR sequences based on a large dataset.

Potential Biases

Potential biases may arise from the selection of training datasets and the inherent limitations of machine learning models.

Limitations

The study's focus on specific LTR sequences may limit the generalizability of the findings to all retrotransposons.

Statistical Information

P-Value

0.09

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/s13040-024-00410-z

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication