Nagarajan Vaidehi

Nagarajan Vaidehi, Ph.D.

  • Chair, Department of Computational and Quantitative Medicine

Nagarajan Vaidehi, Ph.D.

Enfoque de la investigación :
  • Immunology
  • 2005 - present, Professor of Immunology, City of Hope, Duarte, CA
  • 1998 - 2005, Director of Biosimulations, Materials and Process Simulation Center, Caltech, Pasadena, CA
  • 1994 - 1998, Senior Scientist, Materials and Process Simulation Center, Caltech
  • 1991 - 1994, Post-doctoral Fellow at Caltech
  • 1990 - 1991, Post-doctoral Fellow, University of Southern California, Prof. A. Warshel
  • 1989 - 1990, Post-doctoral fellow at University of Exeter, UK


  • 1986, Indian Institute of Technology, Ph.D., Theoretical Chemistry
  • 1981, Indian Institute of Technology, M.Sc., Chemistry (minor- Physics and Mathematics), Highest Honors
Proteins are allosteric nano-machines whose conformational dynamics controls their functional versatility. Conformational dynamics is important in understanding the allosteric nature of proteins, in identifying allosteric druggable sites as well as in designing drugs with functional specificity.
Biophysical experimental methods provide fragmented information on the structure and dynamics and the X-ray crystallography provides a static picture of one of the low energy conformations in an ensemble of states. Therefore computational methods are essential in integrating the experimental information and provide an atomic level detail of the dynamics of proteins. One of the major bottlenecks in using the existing computational methods to study dynamics of proteins is the limitation in time scale and the narrow conformational search afforded by these methods. Thus we need multi-scale computational methods that span a larger range in time and length scale to extend the use of computational methods to large protein complexes. Our laboratory is focused on developing state of the art multi-scale computational methods to study the conformational dynamics of proteins. We are developing coarse grain computational methods to sample the various kinetic states of the protein dynamics, followed by fine grain computational methods to capture the detailed atomic level structural changes and to calculate the thermodynamic properties.
Our research projects include:
  1. Development of constrained molecular dynamics methods – GNEIMO
  2. Development of coarse grained conformational sampling method for G-protein coupled receptors (GPCRs) – GPCRSimKit
  3. Development of computational method for designing thermostable mutants for GPCRs – LITiConDesign
  4. Development and application of computational methods to identify allosteric sites for drug design in protein-protein complexes – AlloBindSite
  5. Application of these methods to design drugs with functional specificity for GPCRs targeting pancreatic cancer and other cancers – Chemokine
A Hierarchical framework for constrained molecular dynamics method
Molecular dynamics (MD) simulation is a powerful computational tool in structural biology, widely used for understanding conformational changes in proteins, and folding of peptides. However MD simulations using Cartesian dynamics model is limited by the total simulation time scale being in tens of nanoseconds for large proteins. Biological processes on the other hand need microseconds of simulation time. We developed the Internal Coordinate Molecular Dynamics (ICMD) algorithms in the early 1990s to enable larger simulation time-steps and they show great promise in long time scale simulations. Despite their promise, ICMD techniques have made little progress due in large part to the additional mathematical complexity of internal coordinate models. As a NIH-NIGMS project and in collaboration with Dr. Abhi Jain at the NASA Jet Propulsion Laboratory at Caltech, we are developing the ICMD methods called Generalized Newton-Euler Inverse Mass Operator (GNEIMO) to enable long time scale and wider conformational search simulations. These simulations have been applied to various biological problems such as
  1. Study of large scale conformational dynamics of proteins wherein we showed the NMR based ensemble of conformations of calmodulin was sampled by GNEIMO method.
  2. Structural refinement of homology models of proteins.
  3. Ab initio folding of simple proteins.
We are now initiating a collaboration with Lawrence Berkeley laboratory to use the GNEIMO method with the program “PHENIX” that to fit models to X-ray crystallography and low resolution electron microscopy measurements.
The GneimoSim software can be downloaded free of cost for academic use from:
A Multiscale computational framework for studying G-protein coupled receptors (GPCRs)
G-protein coupled receptors (GPCRs) play an important role in the physiology and in the pathophysiology of many serious diseases. They form the largest superfamily of drug targets. Since GPCRs are membrane bound and are highly dynamic, obtaining three dimensional structural information for GPCRs is a feat and it requires a confluence of various biophysical techniques that include computational methods. The crystal structure is a snapshot in the conformational ensemble that the receptor samples in the absence of any stimulant. We are developing multiscale simulation method suite, GPCRSimKit, that integrates coarse grain simulation method with fine grain techniques. The GPCRSimKit will enable simulation of the dynamics of GPCR conformational ensemble starting from the inactive crystal structures or refine homology models for drug design. The GPCRSimkit will allow calculation of the modulation of the potential energy landscape by full, partial, and inverse agonists. This platform of computational techniques, will lay a theoretical basis and play a crucial role as more crystal structures of GPCRs get published.
Nagarajan Vaidehi Lab
A computational method for designing thermostable mutants for GPCRs

G-protein coupled receptors are membrane proteins and play an important part in cellular signal transduction. Solving the three dimensional structures of these proteins is critical and is becoming viable lately. However the biggest bottleneck in obtaining sufficient quantities of the pure protein is that GPCRs are conformationally flexible, and hence aggregate at higher concentrations during purification. A solution to this challenge is to derive thermostable mutants of GPCRs that are amenable to purification techniques. However the experiments involved in identifying the residue positions that lead to thermostability as well as the thermostable mutants is both expensive and time consuming. There are about 300 mutations that need to be done just to be able to identify positions that lead to thermostability. Our goal is to develop a fast computational screening method, LITiConDesign to design and thermally stable mutants of several GPCRs. We will target class A GPCRs in their agonist and antagonist bound structures. This project is in collaboration with Dr. Chris Tate (MRC, Cambridge, UK) and Dr. Reinhard Grisshammer (NINDS).

Development of computational method to identify allosteric sites for drug design in protein-protein complexes
G-protein-coupled receptors (GPCRs) are membrane proteins that allosterically transduce the signal of ligand binding in the extracellular (EC) domain to couple to proteins in the intracellular (IC) domain. However, the complete pathway of allosteric communication from the EC to the IC domain, including the role of individual amino acids in the pathway is not known. Using the correlation in torsion angle movements calculated from microseconds-long molecular-dynamics simulations, we have developed a computational analysis method based on graph theory to elucidate the allosteric pathways in GPCRs. This method is generic and applicable to all proteins. In addition, our analysis shows that mutations that affect the ligand efficacy, but not the binding affinity, are located in the allosteric pipelines. This clarifies the role of such mutations, which has hitherto been unexplained. The residues involved in allosteric communication can be used as “allosteric hubs” that modulate the activity of the protein. We use this information on allosteric hub residues to identify druggable allosteric binding sites in proteins. These potential binding sites can be used to screen for small molecules that act as allosteric modulators or inhibitors to protein-protein interactions.
The dynamics of G-protein coupled receptors (GPCRs) and their relevance in drug design G-protein coupled receptors belong to a superfamily of seven helical transmembrane proteins that play a critical role in many physiological processes. They are implicated in the pathology of many diseases such as atherosclerosis, cancer, auto-immune and auto-inflammatory diseases, cancer metastasis, and hence form the biggest class of drug targets. One of the major complexities in drug design for GPCRs, however, is their conformational flexibility. This dynamic flexibility leads to GPCR conformations being in equilibrium between several active and inactive conformational states. Therefore a molecular level understanding of the dynamics is vital to designing functional selective drugs for GPCRs.

Computational methods for studying the dynamics of GPCR conformations: In my laboratory, we have developed novel computational methods to map the potential energy surface and the dynamics of GPCR conformational states and use them for drug design, as seen in the figure, which shows the potential energy surface of an antagonist bound GPCR (right) and the binding site of an antagonist bound to a GPCR used for drug design (left). We have applied these techniques to design drugs for β adrenergic receptors (targets for hypertension and asthma) and chemokine receptors. Using these methods, we also design thermally stable mutant GPCRs for several class A GPCRs that would strongly aid the crystallization of these receptors.
Targeting Chemokine receptors for pancreatic cancer: We are particularly interested in understanding the structural basis of antagonist binding to chemokine receptors. Chemokine receptors belong to class A GPCRs and show versatile function in regulating immune cells. They are also implicated in autoimmune diseases, as well as cancer. Combining computational methods with site directed mutagenesis we have studied antagonist binding for several chemokine receptors such as CCR1, CCR2, CCR3, CCR5, CXCR1, CXCR2, CXCR3, and CXCR4.
Development of Constrained dynamics methods for long time scale simulations: Molecular dynamics simulations involving all atoms is computationally intensive especially for large proteins or protein-protein complexes and therefore poses a bottleneck for realistic biological simulations. We are using algorithms from robotics in collaboration with NASA-JPL to develop constrained dynamics algorithms. In these methods the protein molecule is modeled as a collection of rigid bodies connected by flexible hinges and the equations of motion are solved in internal coordinates. The major advantage of this method is that it allows large conformational search as well as long time scale simulations.

Development of computational methods to identify allosteric sites to disrupt protein-protein interactions: We are developing computational alanine scanning methods.

Nagarajan Vaidehi Lab 2
Allen Mao Ph.D.
Staff Scientist  
[email protected]
Sangbae Lee Ph.D    
Postdoctoral Fellow    
[email protected]  
Vinod Kasam Ph.D
Postdoctoral Fellow
[email protected]
Saugat Kandel
Research Associate
[email protected]
Adrien Larsen M.S
Research Associate II
[email protected]
Hubert Li      
Graduate Student     
[email protected]
Manbir Sandhu
Graduate student
[email protected]
Ir al Inicio