A passionate PhD scientist specializing in theoretical and computational biophysics, I possess a diverse skill set encompassing statistical mechanics, deep learning, drug discovery, and multi-scale simulations. My research focuses on developing advanced algorithms for quantifying protein-drug interactions, Markov-state modeling, deep-learned kinetic models, and simulation-driven drug design. Recognized for my leadership, community-building, innovation, and communication skills, I excel at fostering collaborative environments and advancing scientific knowledge.

I have authored 12 research publications since 2020, including seven first-author papers in peer-reviewed journals. My industry experience includes internships as a Multiscale Modeling Scientist Intern at Genentech (2024), Computer-Aided Drug Design Intern at Janssen Pharmaceutical Companies of Johnson and Johnson (2023), and Research and Early Development Intern at Genentech, Inc. (2022). I have received numerous awards, such as the Merck Research Award for Underrepresented Chemists of Color (2024), Merkin Graduate Fellowship (2023), Distinguished Graduate Student Fellowship (2022), Swiss Government Excellence Scholarship (2017), and Innovation in Science Pursuit for Inspired Research (INSPIRE) Fellowship (2012). My technical expertise includes programming languages like Python, MATLAB, R, and LATEX, and computational software such as Amber, CHARMM, NAMD, Gaussian, and PyMOL. I am proficient in machine learning frameworks like PyTorch, TensorFlow, and Keras, with advanced skills in deep learning methodologies and Markov state models for accelerated drug discovery projects.

I have conducted collaborative research at the Center for Computational Biology at the Flatiron Institute and the Institute of Physics at the University of Freiburg. My commitment to mentorship is evident through my guidance of underrepresented undergraduate students in the UCSD EXperiential Projects for Accelerated Networking and Development program and my role as a Teaching Assistant for multiple undergraduate courses in the Department of Chemistry at UCSD.

As an active representative of international graduate students, I have been involved with the UCSD Chemistry International Friendship Group, the Chemistry Graduate Student Council, and the UCSD Senate for Academic Freedom. I have delivered multiple talks at conferences such as the American Chemical Society meetings and the WESTPA Developer’s Meeting. Dedicated to advancing scientific innovation and fostering an inclusive academic environment, I am committed to accelerating drug discovery efforts and building efficient computational pipelines through excellence in research and collaboration.

CONTRIBUTIONS TO SCIENCE

Early Career (Undergraduate and Master’s Research)

During my undergraduate studies, my research focused on interdisciplinary science, from studying quantum tunneling phenomena in heterocyclic carbenes to calculating hydrogen diffusion rates on noble metal surfaces and conducting large-scale protein simulations. These projects involved rigorous ab initio electronic structure calculations, molecular dynamics simulations, and novel methods development, which provided theoretical rationales for experimental observations and established valuable benchmarks for future research in the field. For my master’s thesis, I employed quantum mechanical calculations to study molecular interactions of aliphatic amines with the Zinc-based metal-organic frameworks (MOFs), where I identified the ground-state dipole moments as the directive force behind the fluorescence “turn-on” mechanism, allowing for a comprehensive understanding of ligand interaction and binding energy orders in the solid, solution, and vapor phases of MOFs. In a separate study, my research explored the adsorptive elimination of glyphosate from aqueous solutions, employing two zirconium-based frameworks, NU-1000 and UiO-67. By conducting ab initio electronic structure calculations, it was determined that NU-1000 exhibited a higher interaction energy with glyphosate, which aligned with its superior adsorption efficacy and recyclability relative to UiO-67. After completing my master’s degree, I worked on developing algorithms in quantum dynamics to study the concerted hydrogen transfer rates in water clusters. In another study, I employed quantum mechanical calculations to support experimental findings of highly selective and sensitive probes for copper ion detection, focusing on the computation of energy gaps between the highest occupied and lowest unoccupied molecular orbitals of the probes to assess their selectivity.

Publications

Halali V. Vishaka, Manav Saxena, H. R. Chandan, Anupam Ojha, and R. Geetha Balakrishna. “Paper based field deployable sensor for naked eye monitoring of copper (II) ions; elucidation of binding mechanism by DFT studies.” Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, Elsevier; 2019; DOI: 10.1016/j.saa.2019.117291.

Asha Pankajakshan, Mekhola Sinha, Anupam Ojha, and Sukhendu Mandal. “Water-stable nanoscale zirconium-based metal–organic frameworks for the effective removal of glyphosate from aqueous media.” ACS Omega, ACS; 2018; DOI: 10.1021/acsomega.8b00921.

Prabu Mani, Anupam Ojha, Vennapusa Sivaranjana Reddy, and Sukhendu Mandal. “Turn-on fluorescence sensing and discriminative detection of aliphatic amines using a 5-fold-interpenetrated coordination polymer.” Inorganic Chemistry, ACS; 2017; DOI: 10.1021/acs.inorgchem.7b00787.

Graduate Career (Ph.D. Research)

During my Ph.D. research, I have addressed the challenge of accurately predicting the kinetic and thermodynamic properties of complex biomolecular systems. I am one of the lead developers of the state-of-the-art programs in our research laboratory, i.e., SEEKR (Simulation Enabled Estimation of Kinetic Rates), which utilizes Markovian milestoning with Voronoi tessellations approach within the multiscale framework of simulations to enhance simulation efficiency and accurately determining receptor-ligand (un)binding kinetic rates, offering significant advantages in drug discovery. My contributions extended to developing the QMrebind (Quantum Mechanical force field reparameterization at the receptor-ligand binding site) method, which significantly improves the precision of force fields by incorporating quantum mechanical methods into the already existing multiscale simulations in the SEEKR program, providing more accurate estimates of drug-target residence times. Furthermore, I developed the hybrid GaMD-WE (Gaussian accelerated molecular dynamics - weighted ensemble) method, integrating GaMD with the weighted ensemble approach to efficiently sample both the thermodynamic and kinetic properties of interest, exhibiting better performance for complex systems than the conventional simulation methods. I recently developed the DeepWEST (Deep learning for the Weighted Ensemble Simulation Toolkit) method, which employed deep-learned Markov state models and weighted ensemble simulations for data-driven conformational sampling. This approach substantially improved the estimation of kinetics and thermodynamics for complex biomolecular systems. My research findings throughout my Ph.D. have been published in ten peer-reviewed journals, underscoring my commitment to advancing the field of theoretical and computational biophysics, with a specific focus on driving progress in drug discovery and research.

QMrebind: Incorporating quantum mechanical force field reparameterization at the ligand binding site for improved drug-target kinetics through milestoning simulations

The QMrebind method improves the accuracy of multiscale milestoning simulations for receptor-ligand unbinding kinetics by quantum mechanically reparameterizing ligand charges at the binding site. By integrating the QMrebind scheme within the milestoning framework, the study successfully achieved precise kinetic rate predictions for a series of Hsp90-inhibitor complexes compared to simulations using generic force fields, underscoring the potential of quantum mechanical refinement in simulation accuracy. As the lead author of this study, I conceptualized integrating the quantum mechanical reparameterization scheme into the multiscale milestoning simulation approach, aiming to refine the kinetic estimates of receptor-ligand interactions, led the QMrebind software development and modeled the Hsp90-inhibitor complexes prior to simulation. I conducted QM/MM multiscale simulations for a series of Hsp90-inhibitor complexes, trajectory visualization, post-simulation analysis, and residence time calculations for a series of Hsp90-inhibitor complexes.

Publications

Anupam Anand Ojha, Lane William Votapka, and Rommie Amaro. “QMrebind: Incorporating quantum mechanical force field reparameterization at the ligand binding site for improved drug-target kinetics through milestoning simulations.” Chemical Science, RSC; 2023; DOI: 10.1039/D3SC04195F.

DeepWEST: Deep Learning of Kinetic Models with the Weighted Ensemble Simulation Toolkit for Enhanced Sampling

The DeepWEST approach is a novel method that combines deep-learned Markov state models (MSMs) with the weighted ensemble (WE) simulations to enhance the sampling of kinetic models in molecular dynamics (MD) simulations. Its novelty lies in utilizing short unbiased MD trajectories and deep learning to identify statistically significant metastable states as starting structures to WE simulations, resulting in a faster and more accurate estimation of kinetic properties with less computational demand than existing methods. This hybrid approach significantly improves the sampling of rare events in biomolecular complexes and offers an automated end-to-end workflow, showcasing superior performance on complex biomolecular systems. As the lead author of this study, I conceptualized integrating the deep-learning algorithms based on the Variational approach for Markov processes (VAMP) with the weighted ensemble simulations for faster kinetic estimates of biomolecular complexes. I led the DeepWEST software development and conducted WE simulations for alanine dipeptide, chignolin, and NTL9 systems, trajectory visualization, post-simulation analysis, and transition rate calculations between the metastable states of these systems.

Publications

Anupam Ojha, Saumya Thakur, Surl-Hee Ahn, and Rommie E. Amaro. “DeepWEST: Deep Learning of Kinetic Models with the Weighted Ensemble Simulation Toolkit for Enhanced Sampling.” Journal of Chemical Theory and Computation, ACS; 2023; DOI: 10.1021/acs.jctc.2c00282.

Selectivity and rank ordering of tight-binding JAK-STAT pathway inhibitors using Markovian milestoning with Voronoi tessellation simulations

A multiscale simulation method, i.e., Markovian milestoning with Voronoi tessellations (MMVT) simulation approach, is employed to enhance the accuracy of measuring both kinetic and thermodynamic aspects of JAK-inhibitor complexes. The MMVT approach, along with long-scale atomistic simulations, provided a detailed examination of the binding mechanisms, potentially guiding the understanding of targeted and efficient therapeutic inhibitors. The study demonstrated the selectivity of JAK2 inhibitors over JAK3, offering promising therapeutic strategies for myeloproliferative disorders. As the lead author of this study, I conceptualized the study of JAK-inhibitor complexes using the multiscale Markovian with Voronoi tessellations (MMVT) approach to determine the residence times of JAK2 and JAK3 inhibitors. This approach was instrumental in elucidating the preferential selectivity of JAK2 inhibitors over JAK3, along with modeling the JAK-inhibitor complexes and conducting multiscale simulations followed by trajectory visualization, post-simulation analysis, and drug-target residence time calculations for JAK-inhibitor complexes. I ran the quantum mechanical calculations for the ligands in this study to demonstrate the effect of specific residue-ligand interactions.

Publications

Anupam Ojha, Ambuj Srivastava, Lane Votapka, and Rommie Amaro. “Selectivity and rank ordering of tight-binding JAK-STAT pathway inhibitors using Markovian milestoning with Voronoi tessellation simulations.” Journal of Chemical Information and Modeling, ACS; 2023; DOI: 10.1021/acs.jcim.2c01589.

Gaussian-Accelerated Molecular Dynamics with the Weighted Ensemble Method: A Hybrid Method Improves Thermodynamic and Kinetic Sampling

GaMD-WE is a novel hybrid sampling method combining Gaussian-accelerated molecular dynamics (GaMD) and weighted ensemble (WE) ensemble simulations for efficient and fast sampling of biomolecules. This method leverages the ability of GaMD to efficiently sample the potential energy landscape by lowering energy barriers, followed by the application of WE to calculate the kinetic properties and transition rates between metastable states. This composite approach has demonstrated superior performance in simultaneously capturing thermodynamic and kinetic properties more accurately than either method could achieve independently. Specifically, our hybrid method has showcased significant advancements in the study of larger systems, with a notable improvement in sampling for the Bovine Pancreatic Trypsin Inhibitor (BPTI) system due to its expansive exploration of the free-energy landscape. As an equal contributing author to this study, I led the software and conceptual development of the GaMD-WE method, modeled the complexes in the study, and calculated the transition rates between the metastable states using the WE method.

Publications

Surl-Hee Ahn, Anupam Ojha, Rommie E. Amaro, and J. Andrew McCammon. “Gaussian-Accelerated Molecular Dynamics with the Weighted Ensemble Method: A Hybrid Method Improves Thermodynamic and Kinetic Sampling.” Journal of Chemical Theory and Computation, ACS; 2021; DOI: 10.1021/acs.jctc.1c00770 (Co-Authors).

Predicting ligand binding kinetics using a Markovian milestoning with Voronoi tessellations multiscale approach

In 2017, we introduced a novel implementation of a Markovian milestoning with Voronoi tessellations approach within the multiscale simulation scheme, i.e., SEEKR (Simulation Enabled Estimation of Kinetic Rates). This method greatly reduces simulation time by up to ten-fold without compromising the accuracy of estimating the kinetic and thermodynamic properties of biomolecular complexes. We showcased the practical utility of this approach in ranking drug-target complexes by kinetic rates and its application to the trypsin-benzamidine complex. Furthermore, we upgraded to SEEKR2, a more versatile and faster iteration of the SEEKR suite with additional capabilities such as hydrogen mass repartitioning. Our study emphasizes the increased speed, improved results, and user-friendly design of the SEEKR2 package, which lays the groundwork for future expansions and applications in molecular kinetics. In these two pivotal studies, where I served as a co-author, I worked on developing and implementing the Markovian scheme with the SEEKR package, modeled the complexes in the study, and ran Brownian dynamics simulations to estimate the binding rates for the ligands. Subsequently, I led the development of a comprehensive tutorial for the SEEKR software program, guiding users through the installation, execution, and analysis of molecular dynamics and Brownian dynamics simulations. This tutorial details the features of SEEKR, such as its compatibility with both NAMD and OpenMM simulation engines and the upgrade to Browndye2 for enhanced simulation efficiency.

Publications

Anupam Ojha, Lane Votapka, Gary Huber, Shang Gao, and Rommie Amaro. “An introductory tutorial to the SEEKR2 (Simulation enabled estimation of kinetic rates v. 2) multiscale milestoning software.” Living Journal of Computational Molecular Science; 2024; DOI: l10.33011/livecoms.5.1.2359.

Lane Votapka, Andrew Stokely, Anupam Ojha, and Rommie Amaro. “SEEKR2: Versatile Multiscale Milestoning Utilizing the OpenMM Molecular Dynamics Engine.” Journal of Chemical Information and Modeling, ACS; 2022; DOI: 10.1021/acs.jcim.2c00501.

Benjamin R. Jagger, Anupam Ojha, and Rommie Amaro. “Predicting ligand binding kinetics using a Markovian milestoning with Voronoi tessellations multiscale approach.” Journal of Chemical Theory and Computation, ACS; 2020; DOI: 10.1021/acs.jctc.0c00495.

Other Noteworthy Publications

Xandra Nuqui, Lorenzo Casalino, Ling Zhou, Mohamed Shehata, Albert Wang, Alexandra L. Tse, Anupam Ojha, Fiona L. Kearns, Mia A. Rosenfeld, Emily Happy Miller, Cory M. Acreman, Surl-Hee Ahn, Kartik Chandran, Jason S. McLellan, Rommie E Amaro. “Simulation-Driven Design of Stabilized SARS-CoV-2 Spike S2 Immunogens.” bioRxiv; 2023; DOI: 10.1101/2023.10.24.563841.

Anisha Nigam, Jeremiah D. Momper, Anupam Ojha, and Sanjay Nigam. “Distinguishing Molecular Properties of OAT, OATP, and MRP Drug Substrates by Machine Learning.” Pharmaceutics, MDPI; 2024; DOI: 10.3390/pharmaceutics16050592.

Anisha K. Nigam, Anupam Ojha, Julia G. Li, Da Shi, Vibha Bhatnagar, Kabir B. Nigam, Ruben Abagyan, and Sanjay K. Nigam. “Molecular Properties of Drugs Handled by Kidney OATs and Liver OATPs Revealed by Chemoinformatics and Machine Learning: Implications for Kidney and Liver Disease.” Pharmaceutics, MDPI; 2021; DOI: 10.3390/pharmaceutics13101720 (Co-Authors).