HBcompare: Classifying Ligand Binding Preferences with Hydrogen Bond Topology.

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Protein structure analysis often uses holistic representations, making it difficult to associate function with specific structural characteristics.
• Understanding the role of specific interactions, like hydrogen bonds, in protein function is crucial for elucidating biochemical mechanisms.

Purpose of the Study:

• To introduce HBcompare, a method for classifying protein structures based solely on hydrogen bond topology.
• To demonstrate HBcompare's ability to implicate hydrogen bonds in selective ligand binding.
• To establish a novel approach for the automatic explanation of biochemical mechanisms.

Main Methods:

• HBcompare analyzes hydrogen bond topology to categorize protein structures by ligand binding preference.
• The method excludes other protein structural characteristics to isolate the role of hydrogen bonds.
• Five datasets of proteins with similar reactions but different ligand preferences were used for evaluation.

Main Results:

• HBcompare correctly categorized proteins with similar ligand binding preferences with 89.5% accuracy.
• Classification accuracy using only hydrogen bond topology surpassed standard structure-based comparison algorithms.
• The method successfully implicated hydrogen bonds in selective ligand binding.

Conclusions:

• HBcompare is the first technique capable of classifying protein structures based solely on hydrogen bond topology.
• This method provides a significant step towards automatically explaining biochemical mechanisms by highlighting the role of hydrogen bonds.
• HBcompare offers a new tool for understanding protein function and ligand specificity.