Comparative footprinting of DNA-binding proteins

Area of Science:

• Computational Biology
• Structural Biology
• Bioinformatics

Background:

• Comparative modeling is a key computational technique in molecular biology and biotechnology.
• While traditionally used for protein structures and protein-ligand interactions, its application has expanded to protein-protein interfaces.
• This research pioneers the systematic analysis of comparative models for protein-DNA complexes to predict DNA binding sites.

Purpose of the Study:

• To systematically analyze the feasibility and utility of comparative models for protein-DNA complexes.
• To assess the potential of these models in predicting DNA binding sites.
• To develop and benchmark a protocol for comparative modeling of protein-DNA interactions.

Main Methods:

• Analysis of structural and evolutionary conservation patterns at protein-DNA interfaces.
• Modeling of contacting residues and their side chains to identify nucleotide interactions.
• Implementation and benchmarking of the DNASITE protocol using a dataset of 85 regulators from Escherichia coli.

Main Results:

• Structural and evolutionary conservation at protein-DNA interfaces impose limitations on modeling accuracy.
• Side chains of contacting residues can be reliably modeled, enabling the identification of contacting nucleotides.
• The DNASITE protocol demonstrated that comparative footprinting can generate useful predictions based on structural data, with accuracy dependent on interface similarity to the template.

Conclusions:

• Comparative modeling is a viable approach for studying protein-DNA complexes.
• The DNASITE protocol provides a useful tool for predicting DNA binding sites using structural information.
• The accuracy of comparative footprinting is influenced by the conservation of the protein-DNA interface relative to the modeling template.