TY - JOUR
T1 - Genome-wide computational determination of the human metalloproteome
AU - Azia, Ariel
AU - Levy, Ronen
AU - Unger, Ron
AU - Edelman, Marvin
AU - Sobolev, Vladimir
N1 - Publisher Copyright:
© 2015 Wiley Periodicals, Inc.
PY - 2015/5/1
Y1 - 2015/5/1
N2 - Accurate prediction of protein function in humans is important for understanding biological processes at the molecular level in biomedicine and drug design. Over a third of proteins are commonly held to bind metal, and ∼10% of human proteins, to bind zinc. Therefore, an initial step in protein function prediction frequently involves predicting metal ion binding. In recent years, methods have been developed to predict a set of residues in 3D space forming the metal-ion binding site, often with a high degree of accuracy. Here, using extensions of these methods, we provide an extensive list of human proteins and their putative metal ion binding site residues, using translated gene sequences derived from the complete, resolved human genome. Under conditions of ∼90% selectivity, over 900 new human putative metal ion binding proteins are identified. A statistical analysis of resolved metal ion binding sites in the human metalloproteome is furnished and the importance of remote homology analysis is demonstrated. As an example, a novel metal-ion binding site involving a complex of a botulinum substrate with its inhibitor is presented. On the basis of the location of the predicted site and the interactions of the contacting residues at the complex interface, we postulate that metal ion binding in this region could influence complex formation and, consequently, the functioning of the protein. Thus, this work provides testable hypotheses about novel functions of known proteins.
AB - Accurate prediction of protein function in humans is important for understanding biological processes at the molecular level in biomedicine and drug design. Over a third of proteins are commonly held to bind metal, and ∼10% of human proteins, to bind zinc. Therefore, an initial step in protein function prediction frequently involves predicting metal ion binding. In recent years, methods have been developed to predict a set of residues in 3D space forming the metal-ion binding site, often with a high degree of accuracy. Here, using extensions of these methods, we provide an extensive list of human proteins and their putative metal ion binding site residues, using translated gene sequences derived from the complete, resolved human genome. Under conditions of ∼90% selectivity, over 900 new human putative metal ion binding proteins are identified. A statistical analysis of resolved metal ion binding sites in the human metalloproteome is furnished and the importance of remote homology analysis is demonstrated. As an example, a novel metal-ion binding site involving a complex of a botulinum substrate with its inhibitor is presented. On the basis of the location of the predicted site and the interactions of the contacting residues at the complex interface, we postulate that metal ion binding in this region could influence complex formation and, consequently, the functioning of the protein. Thus, this work provides testable hypotheses about novel functions of known proteins.
KW - Botulinum substrate
KW - Human proteome
KW - Metal binding sites
KW - Protein function prediction
UR - http://www.scopus.com/inward/record.url?scp=84927690093&partnerID=8YFLogxK
U2 - 10.1002/prot.24790
DO - 10.1002/prot.24790
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
C2 - 25739467
AN - SCOPUS:84927690093
SN - 0887-3585
VL - 83
SP - 931
EP - 939
JO - Proteins: Structure, Function and Bioinformatics
JF - Proteins: Structure, Function and Bioinformatics
IS - 5
ER -