Associate Professor
Department of Chemical & Biomolecular Engineering
Whiting School of Engineering

jgray@jhu.edu

208 Maryland Hall
3400 N. Charles Street
Baltimore, MD 21218

Office: 410-516-5313
Lab: 410-516-7308

Proteins self-assemble into intricate and functional structures.  Protein structure determines function, and knowledge of the three-dimensional molecular structure of proteins and their complexes is essential to understand biological mechanisms and to rationally engineer proteins.  The standard methods for determining structure are x-ray crystallography and nuclear magnetic resonance spectroscopy.  Determining protein structure by these techniques can be difficult or even impossible, and it always is expensive.  Computational approaches provide a faster and cheaper alternate route to determine structures that are difficult to obtain otherwise. 

The goal of our research is to create and apply protein structure prediction methods to address practical problems in self-assembly and function in biomolecular engineering.  We focus in the areas of protein-protein docking, therapeutic antibodies, allostery, and protein-surface interactions.  These problems currently are at or beyond the limits of resolution and complexity that may be addressed computationally.  Further, they are relevant for practical purposes in biology and biomolecular engineering.  For example, antibodies are the fastest growing class of drugs, proteins interacting with solid surfaces are central in bionanotechnology and biomineralization, and allostery plays a critical role in cellular regulation and can be exploited for nano/molecular switches.

Computational prediction of protein structure is an important but difficult problem limited by two fundamental challenges.  First, the conformation space is immense: the famous Levinthal paradox estimates a 100-amino acid protein as having 3¹⁰⁰ is approximately equal to 10⁴⁸ conformations.  It is not possible to exhaustively sample all of these in search of the native conformation, so strategies are needed to explore the most relevant or realistic conformations efficiently.  Even if all conformations could be sampled, the second challenge is that an accurate energy function is needed to identify the lowest free energy state (assuming the protein reaches equilibrium).  Quantum calculations are infeasible, and empirical energy functions are approximations.  Progress has been achieved on both challenges through advances such as the use of protein structural libraries, rapid discrete and continuous optimization methods, and the use of known protein structures to refine statistical potentials and validate algorithms.  The field of protein structure prediction has reached the point where we can now fold small domains, dock single protein domains, and even design novel folds. 

We are working at the next level beyond the folding of small domains and docking of small rigid domains by adapting techniques developed for model biomolecules and creating new methods to handle the extra complexity that is needed for real-world biological and engineering problems.  Our achievements have included a new and effective flexible-backbone docking method, high-resolution antibody modeling methods, applications of antibodies against anthrax and cancer, the first benchmark set of allosteric structures, network models of allosteric communication, and the first structure-prediction-based method for proteins on solid surfaces.  Ongoing projects include docking methods to accommodate large conformational flexibility, analysis and design of allosteric proteins, and improved methods for capturing the interactions of proteins with surfaces.

Selected Publications
Masica, D.L., and J.J. Gray. (2009) Solution- and adsorbed-state structural ensembles predicted for the statherin-hydroxyapatite system. Biophys. J. 96:3082-3091. *Online*

Daily, M.D., and J.J. Gray. (2009) Allosteric communication occurs via networks of tertiary and quaternary motions in proteins. PLoS Comput. Biol. 5:e1000293. *Online*

Sivasubramanian*, A., Sircar*, A., S. Chaudhury, and J.J. Gray. (2009) Toward high-resolution homology modeling of antibody Fv regions and application to antibody-antigen docking. Proteins: Struct. Funct. Bioinf. 74:497-514. (*These authors contributed equally to this work) *Online*

Chaudhury, S., and J.J. Gray. (2008) Conformer selection and induced fit in flexible backbone protein-protein docking using computational and NMR ensembles. J. Mol. Biol. 381:1068-1087. *Online*

Berrondo, M., M. Ostermeier, and J.J. Gray. (2008) Structure prediction of domain insertion proteins from structures of the individual domains. Structure 16:513-527. *Online*

Daily, M.D., T.J. Upadhyaya, and J.J. Gray. (2008) Contact rearrangements form coupled networks from local motions in allosteric proteins. Proteins 71:455-466.  *Online*

Chaudhury*, S., Sircar*, A., A. Sivasubramanian, M. Berrondo, and J. J. Gray. (2007) Incorporating biochemical information and backbone flexibility in RosettaDock for CAPRI rounds 6-12. Proteins: Struct. Funct. Bioinf. 69:793-800. (*These authors contributed equally to this work.) *Online*