Predicting enhancer-promoter interactions using a stacking-based ensemble strategy.

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Enhancer-promoter interactions (EPIs) are crucial for gene regulation and implicated in disease.
• High-throughput experimental identification of EPIs is costly and time-consuming.
• Existing computational methods struggle to integrate diverse cell line data for EPI prediction.

Purpose of the Study:

• To develop an efficient and accurate computational framework for predicting enhancer-promoter interactions.
• To overcome the limitations of integrating heterogeneous feature representations from multiple cell lines.
• To enhance the understanding of transcriptional regulation mechanisms.

Main Methods:

• A stacked ensemble framework was proposed, integrating multiple machine learning algorithms.
• Feature representations from diverse cell line datasets were utilized.
• An XGBoost classifier was employed to combine complementary patterns and improve robustness.

Main Results:

• The proposed method achieved superior accuracy and generalization compared to existing models.
• An average Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.909 was obtained.
• The framework demonstrated computational efficiency.

Conclusions:

• The stacked ensemble framework provides an effective approach for predicting enhancer-promoter interactions.
• This method offers a robust and efficient solution for analyzing gene regulation data.
• The findings contribute to advancing our understanding of gene regulation and disease mechanisms.