Sequence features that drive human promoter function and tissue specificity.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Promoters are crucial DNA sequences regulating gene transcription initiation.
• Understanding promoter function is key to deciphering gene regulation.
• Large-scale functional characterization of human promoters is needed.

Purpose of the Study:

• To functionally characterize 4575 human promoters across eight cell lines.
• To correlate promoter activity with endogenous gene expression.
• To predict promoter activity using sequence features alone.

Main Methods:

• Transient transfection reporter assays to measure promoter activity.
• RNA sequencing to measure endogenous gene expression.
• Computational modeling using transcription factor binding motifs.

Main Results:

• Significant correlation (r=0.43) between promoter activity and gene expression.
• CG dinucleotide content distinguishes ubiquitous (high CG) from cell-specific (low CG) promoters.
• Computational models achieved 91% AUC, outperforming CG content prediction by 23% and identifying new TF-promoter associations.

Conclusions:

• Sequence features, particularly TF binding potential, are powerful predictors of promoter activity.
• CG content is a useful but limited feature for classifying promoter activity.
• Systematic functional characterization reveals novel insights into gene regulation.