i) Identification of stabilizing interactions (like CH-pi, CH-O, electrophile-nucleophile, aromatic-aromatic etc.) and their implications in protein structures and function
ii) Analysis of protein conformation
iii) Protein folding, threading and prediction of structures
iv) Molecular modelling and dynamics
v) Molecular recognition, protein-protein complexation and ion-binding by proteins
vi) Crystallography of proteins from phage lambda and Vibrio cholerae
vii) Molecular design and docking
viii) Bioinformatics and proteomics
ix) Crystal packing and supramolecular assembly
x) Characterization of the interaction of nanoparticles with globular proteins

Finding patterns in sequence and structure and deriving the underlying factors stabilizing the native fold of proteins and protein-protein recognition have been the hallmark of Prof. Chakrabarti’s work. Following the tradition of GN Ramachandran he has made distinctive contributions in the areas of biomolecular conformation. Structural information stored in the Protein Data Bank has been used in a novel way to derive thermodynamic properties, such as the conformational entropy of folding and a mechanism of helix nucleation. Another topic has been the elucidation of the geometry of binding of ions and the preferential mode of interaction between planar residues in protein structures. Specific non-bonded interactions, such as electrophile-nucleophile, C-H...pi, C-H...O have been identified, which have role in the stability, as well as the function of protein molecules. To understand molecular recognition and specificity the interfaces in protein-protein assemblies have been dissected in a unique way that led to the delineation of a core region, which is occupied by residues that are highly conserved among homologous proteins. The results elucidating the physicochemical features of protein-protein/DNA interfaces can be used in identifying binding patches on the protein surface and would be useful in understanding protein networks, developing docking algorithm and drug design. An empirical algorithm has been developed that can predict the free energy of binding contributed by a given residue located in protein-protein interfaces. He has been involved in the crystallographic analysis of a number of interesting proteins and small molecules; a recent work on the structure of CII from bacteriophage lambda explains why this transcription activator binds a direct repeat DNA sequence. His group has been involved in the development of many web-based bioinformatics tools, which are available in http://www.boseinst.ernet.in/resources/bioinfo/stag.html. His on-going work involving the antimicrobial activity of ZnO nanoparticles and pH-dependent release of drug molecules from gold nanoparticle-conjugated drug composites is likely to find niche applications in pharmaceutical industry.