Predicting transcription factor binding motifs from DNA-binding domains, chromatin accessibility and gene expression...

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Transcription factors (TFs) are critical regulators of gene expression, binding to specific DNA sequences.
• While high-throughput SELEX sequencing has identified motifs for ~400 human TFs, ~800 TFs still lack experimentally derived binding motifs.
• Identifying motifs for these TFs is essential for a comprehensive understanding of transcriptional regulation.

Purpose of the Study:

• To develop a computational method for associating known DNA sequence motifs with transcription factors (TFs) that lack experimentally determined motifs.
• To computationally infer both monomeric and homodimeric binding motifs for TFs.
• To increase the number of human TFs with known binding motifs, thereby enhancing the study of gene regulation.

Main Methods:

• Developed a probabilistic computational framework to associate known motifs with TFs.
• Integrated data on DNA-binding domains, TF specificities, open chromatin, gene expression, and genomic data.
• The method, implemented in MATLAB, infers monomeric and homodimeric binding motifs.

Main Results:

• Successfully assigned DNA binding motifs to 200 transcription factors (TFs) that previously lacked SELEX-derived motifs.
• This represents approximately a 50% increase in motif coverage for TFs compared to existing data.
• The computational approach effectively leverages diverse biological datasets to predict TF binding preferences.

Conclusions:

• The presented computational method significantly expands the repertoire of known human TF binding motifs.
• This approach reduces the need for extensive experimental efforts in motif discovery.
• The findings provide deeper insights into transcriptional regulatory networks and gene expression control.