AttnW2V-Enhancer: Leveraging attention and Word2Vec for enhanced enhancer prediction

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Enhancer regions are critical regulatory elements influencing gene expression.
• Identifying enhancers is challenging due to genomic sequence complexity and variability.
• Current methods for enhancer detection have limitations in efficiency and interpretability.

Purpose of the Study:

• To develop a novel computational model for accurate enhancer identification.
• To improve upon existing methods for predicting regulatory DNA sequences.
• To provide a more efficient and interpretable framework for enhancer prediction.

Main Methods:

• Proposed AttnW2V-Enhancer model integrating Word2Vec sequence encoding, Convolutional Neural Networks (CNNs), and attention mechanisms.
• Utilized Word2Vec embeddings to capture biologically meaningful patterns in DNA sequences.
• Employed attention mechanisms to dynamically focus on relevant sequence regions for enhanced feature learning.

Main Results:

• AttnW2V-Enhancer achieved high performance on an independent test set: 81.75% accuracy, 83.50% sensitivity, 80.00% specificity, and a Matthews Correlation Coefficient (MCC) of 0.635.
• The model outperformed existing methods for enhancer prediction.
• Demonstrated the effectiveness of attention mechanisms in improving feature learning and model interpretability.

Conclusions:

• The integration of Word2Vec encoding with CNNs and attention mechanisms provides a powerful and interpretable framework for enhancer prediction.
• AttnW2V-Enhancer offers valuable insights for identifying regulatory DNA sequences.
• The developed model represents a significant advancement in the field of computational genomics and gene regulation analysis.