1. Designed graph convolution neural network and transfer learning-based models for screening ultra-large chemical libraries to discover novel therapeutics and predict their toxicity against military-relevant targets.2. Investigated the application of recurrent neural networks and transformers to improve model performance on imbalanced public bioactivity datasets.

1. Designed a comprehensive molecular dynamics (MD) simulations based protocol to elucidate the role of intramolecular interactions and secondary structures in determining sequence-dependent backbone dynamics of disordered proteins. 2. Decoded and characterized the structure, dynamics, binding, and membrane association of FtsQ and SepF, two Mycobacterium tuberculosis drug targets, in collaboration with experimental researchers. Achieved high accuracy in alignment with experimental results, providing deeper insights into these proteins' roles in cell division.3. Identified the mechanism of synergistic activation and allosteric regulation of the intrinsically disordered WASP and N-WASP proteins in response to lipids and proteins. Estimated the free energy of activation using umbrella sampling and the string method.

Taught General Chemistry I and II (Fall 2018, Spring 2019, Summer 2019 and Fall 2019). Responsibilities included Leading laboratory sessions and discussion classes, Grading and Holding office hours

Screened over 1 million small molecules against various neurological targets using ADME filtering and docking techniques. Utilized pharmacophore modeling to identify binding sites and performed MM-GBSA calculations to rank new compounds.

Conducted hybrid QM/MM (Quantum Mechanics / Molecular Mechanics) simulations using the replica-exchange string method to identify the mechanisms of ATP hydrolysis in kinesin motor proteins. Computed the free energy of hydrolysis utilizing umbrella sampling to rank the most viable options.

Molecular Dynamics simulations to understand DNA damage and repair in catalyzed by a photolyase.

Molecular Dynamics simulations to elucidate unfolding of an S-Peptide analogue in presence of urea and attenuation of this effect by introducing an ammonium based ionic liquid along with parametrization of these non-standard residues using Gaussian

Used density functional theory to compare binding poses, interaction energies and density of states of the four DNA Nucleobases on two-dimensional Molybdenum Disulphide and Carbon Nitride sheets.