Similarity regression predicts evolution of transcription factor sequence specificity

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Transcription factor (TF) binding specificities (motifs) are crucial for understanding gene regulation.
• Accurate TF motif prediction is challenging due to the vast number of TFs and genomes.
• Previous studies on motif conservation across species have been limited by prediction method uncertainties.

Purpose of the Study:

• To develop an improved method for predicting TF binding motifs.
• To update and expand the Cis-BP database with more accurate motif predictions.
• To investigate the extent of TF motif evolution and diversification across eukaryotic species.

Main Methods:

• Development and application of a novel method termed similarity regression for TF motif prediction.
• Updating and expanding the Cis-BP database using the new prediction method.
• Comparative analysis of TF motifs between human and Drosophila to assess conservation and divergence.

Main Results:

• Similarity regression significantly improves TF motif prediction accuracy.
• The updated Cis-BP database provides a more comprehensive resource for TF binding motifs.
• Analysis revealed substantial divergence in TF motifs between human and Drosophila, with nearly half of motifs unique to each species.
• Extensive divergence in C2H2 zinc finger proteins was identified as a major driver of motif diversification.

Conclusions:

• TF motif diversification is pervasive across eukaryotes, contrary to previous assumptions of high conservation.
• The new similarity regression method and updated Cis-BP database offer valuable tools for studying TF diversity and gene regulation.
• Understanding TF motif evolution is critical for accurate gene regulation analysis in comparative genomics.