Narendra Kumar

Postdoctoral Fellow

P1100529-send.jpg

Ph.D., National Institute of Immunology, New Delhi, 2009

Center for the Study of Systems Biology
250 14th Street, NW, Room 110
Atlanta GA, 30318
Tel: (404) 407-8976
Fax: (404) 385-7478
narendra.kumar@biology.gatech.edu


Research Interests

I am interested in developing novel computational methods for exploring the changes in small molecule metabolic profile during the disease state, to gain insight into its cause and the effect on metabolic network in pathological state. The major focus of my research is cancer metabolomics and identification of new drug targets.


Publications

Zhou, H, Gao M, Kumar N, Skolnick J.  Submitted.  SUNPRO: Structure and function predictions of proteins from representative organisms. BMC Bioinformatics.
Kumar, N, Skolnick J.  Submitted.  EFICAz2.5: Application of a high-precision enzyme function predictor to 373 proteomes. BMC Bioinformatics.
Anand, S, Prasad MVR, Yadav G, Kumar N, Shehara J, Ansari MZ, Mohanty D.  2010.  SBSPKS: structure based sequence analysis of polyketide synthases. Nucleic Acids Research. 38(Web Server):W487-W496.
Kumar, N, Mohanty D.  2010.  Identification of substrates for Ser/Thr kinases using residue-based statistical pair potentials.. Bioinformatics (Oxford, England). 26(2):189-97. Abstract
Kumar, N, Mohanty D.  2007.  MODPROPEP: a program for knowledge-based modeling of protein-peptide complexes. Nucleic acids research. 35(Web Server issue):W549-55. Abstract

Skolnick Group

Services
PSIFR
Protein Structure Prediction
chunk-TASSER
MetaTASSER
pro-sp3-TASSER
TASSER
Protein Structure Refinement
BSR
Protein Structure Alignment
Fr-TM-align
iAlign
DNA-binding Prediction
DBD-Threader
DBD-Hunter
DP-dock
Protein Function Prediction
FINDSITE
FINDSITE-metal
EFICAz2.5
Ligand Docking/Screening
FINDSITELHM
Q-DockLHM
Software
software-icon.png
EFICAz2.5
FINDSITE
FINDSITELHM
FINDSITE-metal
Fr-TM-align
iAlign
PULCHRA
TASSER-Lite
Databases
Structure Prediction Libraries
gpcr_0.jpg
Human GPCRs
E. coli Proteome
PDB-like Structures
Function Prediction Libraries
kinase1.jpg
KinomeLHM
X-ReactKIN
Human Proteome
Human DNA Binders
Reference Libraries
3jxv_bio_r_500.jpg
Active Site Templates
Apo/Holo Pairs
Simulations
molecular_system.png
E. coli Intracellular Dynamics
PSIFR - a web portal that provides access to TASSER, TASSER-Lite and MetaTASSER and DBD-Hunter, and the enzyme function inference engine EFICaz2.
MetaTASSER - a protein structure prediction server that uses three
state-of-the-art threading methods PROSPECTOR_3, SP3 and SPARKS to identify templates that are selected by 3D-Jury. For targets with no reliable templates, chunk models from chunk-TASSER are also included. SPICKER clustering is used to select the final models from the TASSER trajectory. Detailed atomic models are then built using PULCHRA.
pro-sp3-TASSER - uses a single threading method with multiple scoring to identify templates. Short TASSER runs generate full length models that are selected by TASSER-QA and FTCOM ranking procedures. pro-sp3-TASSER performs better than MetaTASSER for medium/hard targets but is computationally more expensive.
Fr-TM-align - structural alignment program for monomeric proteins that uses the TM-score as the structure comparison metric.
iAlign - structural alignment algorithm for protein-protein interfaces that uses the TM-score as well as a side-chain contact overlap weighted TM-score as the interface comparison metric.
DBD-Threader - a server that predicts DNA-binding protein domains from a protein's sequence.
DBD-Hunter - a server that predicts DNA-binding protein domains given a protein’s structure.
DP-dock - a server that predicts the structure of the DNA/protein complex structure given the tertiary structure of a known DNA-binding protein.
FINDSITE - a threading-based binding site prediction/protein functional inference/ligand screening algorithm that detects common ligand binding sites in a set of evolutionarily related proteins. Crystal structures as well as protein models can be used as the target structures.
FINDSITE-metal - a simple extension of FINDSITE that predicts metal-binding sites, residues and binding metal preferences from evolutionarily related templates detected by threading. Furthermore, it employs machine learning to significantly improve the prediction accuracy particularly against modeled protein structures.
FINDSITELHM - a homology modeling approach to flexible ligand docking. It uses a collection of common molecule substructures derived from evolutionarily related templates as the reference compounds in similarity-based ligand binding pose prediction.
Q-DockLHM - a low-resolution ligand docking/refinement approach applicable to the modeled structures of target receptors. It uses pocket-specific statistical potentials and harmonic restraints imposed on the binding poses of the common molecule substructures extracted from evolutionarily related proteins.
EFICAz2.5 - an accurate sequence based approach to enzyme function inference.
chunk-TASSER - a protein structure prediction method that combines threading templates from SP3 and ab initio folded chunk structures (three consecutive segments of regular secondary structures).
BSR - Binding Site Refinement employs a new template-based method for the local refinement of ligand-binding regions in protein models using closely as well as distantly related templates identified by threading.
FINDSITE - a threading-based binding site prediction/protein functional inference/ligand screening algorithm that detects common ligand binding sites in a set of evolutionarily related proteins.
Fr-TM-align - a program for protein structural alignment that uses the TM-score as the structural comparison metric. This is an improved version of Tm-align.
iAlign - a structural alignment algorithm for the protein-protein interface that uses the TM-score as well as a side-chain contact overlap weighted TM-score as the interface comparison metric.
PULCHRA - a program for all-atom reconstruction from the backbone C-alpha atoms.
TASSER-Lite - a comparative protein structure modeling tool.
Human GPCRs - Predicted structures of 907 G protein-coupled receptors in the Human proteome.
Human DNA Binders - DBD-Threader Predictions on the Human proteome. A list of predicted Human DNA-binding proteins their Sequences and predicted 3D models.
E. coli Poteome - Predicted structures of the E. Coli proteome.
PDB-like Structures - A library of 15,000 compact homopolypeptide structures that cover the PDB.
Active Site Templates - Automated Functional Templates (AFT) for Enzyme Active Site Prediction.
Apo/Holo Pairs - Apo/Holo and Holo/Holo protein pairs. The global structures of apo and holo protein.
KinomeLHM - Comprehensive structural and functional characterization of the human kinome by protein structure predication and ligand virtual screening.
Human Proteome - Structure modeling, function annotation and ligand virtual screening of the human proteome.
TASSER - A meta server for protein structure prediction that combines chunk-TASSER, pro-sp3-TASSER and MetaTASSER.
FINDSITELHM - A homology modeling approach to flexible ligand docking. It uses a collection of common molecule substructures derived from evolutionarily related templates as the reference compounds in similarity-based ligand binding pose prediction.
E. coli Intracellular Dynamics - The inside of cells is highly crowded with bio-macromolecules, such as proteins, DNA, RNA. This environment is quite different from the conditions in a test tube that we usually use for analysis of biomolecular functions, where the macromolecular concentration is ~100 times less than inside cells.
X-ReactKIN - Virtual cross-reactivity profiles of the human kinome.