EMERGE: a flexible modelling framework to predict genomic regulatory elements from genomic signatures

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Regulatory DNA elements control gene expression and are linked to diseases.
• Functional validation of these elements is challenging due to low success rates and vast, growing datasets.
• Current methods struggle to efficiently integrate diverse genomic data for accurate prediction.

Purpose of the Study:

• To develop a user-friendly tool for integrating diverse genomic datasets.
• To enhance the accuracy and success rate of predicting functional regulatory DNA elements.
• To provide a flexible framework for analyzing and visualizing regulatory element data.

Main Methods:

• Developed the EMERGE program for merging and analyzing genomic datasets (e.g., ATAC-seq, ChIP-seq, conservation).
• Utilized a logistic regression framework with optimal dataset weighting based on validated functional elements.
• Employed Receiver Operating Characteristic (ROC) curve analysis to assess prediction performance.

Main Results:

• EMERGE successfully integrates information from multiple genomic datasets.
• Combined datasets significantly improve the prediction of tissue-specific enhancers across species (human, mouse, Drosophila).
• The integrated approach is expected to substantially increase the success rates of functional assays.

Conclusions:

• The EMERGE program offers a powerful, flexible tool for regulatory DNA element prediction.
• Integrating diverse genomic data enhances the identification of functional elements.
• This approach facilitates more efficient and successful functional validation, advancing disease research.