Hydrogen bonds play a vital role in protein-DNA interactions. In particular, side chain-base hydrogen bonds are crucial to the binding specificity between protein and DNA. Mutations effecting interface hydrogen bonds in protein-DNA complexes have been linked to changes in binding specificity and are implicated in various diseases. However, knowledge about the distribution of hydrogen bond energy (HBE) in protein-DNA complexes as compared to other important biomolecular complexes is unknown. Here, we performed a systematic comparative analysis of hydrogen bond energy (HBE) in three protein-ligand complexes; protein-DNA, protein-protein and protein-peptide. Our results show that while the hydrogen bonds in protein-protein and protein-peptide complexes are predominantly strong, a unique, almost equal distribution of strong and weak hydrogen bonds is observed in protein-DNA complexes. More importantly, more strong hydrogen bonds are observed in the minor grooves of highly specific protein-DNA complexes than multispecific complexes indicating the role of minor groove hydrogen bonds in protein-DNA binding specificity. The knowledge gained from these analyses was applied to develop a novel hydrogen bond energy-based method to assess the similarity between protein-DNA complex models and reference structures, an important step towards computational prediction of complex structures. We show that HBE based method provides more accurate assessment of similarity for models generated by both homology modeling and computational docking methods.