A5067081047- Protein Structure and Dynamics
- Gene Regulatory Network Analysis
- SARS-CoV-2 and COVID-19 Research
- Machine Learning in Materials Science
- RNA and protein synthesis mechanisms
- Lipid Membrane Structure and Behavior
- Molecular Junctions and Nanostructures
- Spectroscopy and Quantum Chemical Studies
- Influenza Virus Research Studies
- Mass Spectrometry Techniques and Applications
- Chemical Synthesis and Analysis
- Microbial Metabolic Engineering and Bioproduction
- Various Chemistry Research Topics
- Bioinformatics and Genomic Networks
- Enzyme Structure and Function
- Fuel Cells and Related Materials
- Complex Network Analysis Techniques
- Drug Transport and Resistance Mechanisms
- Lanthanide and Transition Metal Complexes
- Machine Learning in Bioinformatics
- Single-cell and spatial transcriptomics
- Simulation Techniques and Applications
- Bacteriophages and microbial interactions
- COVID-19 epidemiological studies
- DNA and Nucleic Acid Chemistry
University of Pittsburgh
2018-2025
Stony Brook University
2024-2025
Stony Brook School
2025
University of Washington
2021
Massachusetts Institute of Technology
2021
SARS-CoV-2 infection is controlled by the opening of spike protein receptor binding domain (RBD), which transitions from a glycan-shielded 'down' to an exposed 'up' state bind human angiotensin-converting enzyme 2 and infect cells. While snapshots states have been obtained cryo-electron microscopy tomagraphy, details RBD-opening transition evade experimental characterization. Here over 130 µs weighted ensemble simulations fully glycosylated ectodomain allow us characterize more than 300...
We develop a generalizable AI-driven workflow that leverages heterogeneous HPC resources to explore the time-dependent dynamics of molecular systems. use this investigate mechanisms infectivity SARS-CoV-2 spike protein, main viral infection machinery. Our enables more efficient investigation in variety complex environments, including within complete envelope simulation, which contains 305 million atoms and shows strong scaling on ORNL Summit using NAMD. present several novel scientific...
The weighted ensemble (WE) family of methods is one several statistical mechanics-based path sampling strategies that can provide estimates key observables (rate constants and pathways) using a fraction the time required by direct simulation such as molecular dynamics or discrete-state stochastic algorithms. WE oversee numerous parallel trajectories intermittent overhead operations at fixed intervals, enabling facile interoperability with any engine. Here, we report on major upgrades to...
We seek to completely revise current models of airborne transmission respiratory viruses by providing never-before-seen atomic-level views the SARS-CoV-2 virus within a aerosol. Our work dramatically extends capabilities multiscale computational microscopy address significant gaps that exist in experimental methods, which are limited their ability interrogate aerosols at atomic/molecular level and thus obscure our understanding transmission. demonstrate how integrated data-driven platform...
Abstract SARS-CoV-2 infection is controlled by the opening of spike protein receptor binding domain (RBD), which transitions from a glycan-shielded “down” to an exposed “up” state in order bind human ACE2 and infect cells. While snapshots states have been obtained cryoEM cryoET, details RBD transition evade experimental characterization. Here, over 130 μs weighted ensemble (WE) simulations fully glycosylated ectodomain allow us characterize more than 300 continuous, kinetically unbiased...
A promising approach for simulating rare events with rigorous kinetics is the weighted ensemble path sampling strategy. One challenge of this strategy division configurational space into bins sampling. Here we present a minimal adaptive binning (MAB) scheme automated, placement along progress coordinate within framework Results reveal that MAB scheme, despite its simplicity, more efficient than manual, fixed in generating transitions over large free energy barriers, diversity pathways,...
We develop a generalizable AI-driven workflow that leverages heterogeneous HPC resources to explore the time-dependent dynamics of molecular systems. use this investigate mechanisms infectivity SARS-CoV-2 spike protein, main viral infection machinery. Our enables more efficient investigation in variety complex environments, including within complete envelope simulation, which contains 305 million atoms and shows strong scaling on ORNL Summit using NAMD. present several novel scientific...
Passive permeability of a drug-like molecule is critical property assayed early in drug discovery campaign that informs medicinal chemist how well compound can traverse biological membranes, such as gastrointestinal epithelial or restrictive organ barriers, so it perform specific therapeutic function. However, the challenge remains development method, experimental computational, which both determine permeation rate and provide mechanistic insights into transport process to help with rational...
The pathways by which a molecular process transitions to target state are highly sought-after as direct views of transition mechanism. While great strides have been made in the physics-based simulation such pathways, analysis these can be major challenge due their diversity and variable lengths. Here, we present LPATH Python tool, implements semiautomated method for linguistics-assisted clustering into distinct classes (or routes). This involves three steps: 1) discretizing configurational...
The weighted ensemble (WE) strategy has been demonstrated to be highly efficient in generating pathways and rate constants for rare events such as protein folding binding using atomistic molecular dynamics simulations. Here we present five tutorials instructing users the best practices preparing, carrying out, analyzing WE simulations various applications WESTPA software. Users are expected already have significant experience with running standard underlying engine of interest (e.g. Amber,...
Understanding the effects of missense mutations or single amino acid variants (SAVs) on protein function is crucial for elucidating molecular basis diseases/disorders and designing rational therapies. We introduce here Rhapsody-2, a machine learning tool discriminating pathogenic neutral SAVs, significantly expanding precursor limited by availability structural data. With advent AlphaFold2 as powerful structure prediction, Rhapsody-2 trained expanded dataset 117,525 SAVs corresponding to...
The SARS-CoV-2 spike glycoprotein binds human epithelial cells and enables infection through a key conformational transition that exposes its receptor binding domain (RBD). Experimental evidence indicates mutations, particularly the early D614G variant, alter rate of this shift, potentially increasing viral infectivity. To investigate how mutations reshape landscape, we conducted extensive weighted ensemble simulations Ancestral, Delta, Omicron BA.1 strains along RBD opening pathway. We...
A major challenge for many rare-event sampling strategies is the identification of progress coordinates that capture slowest relevant motions. Machine-learning methods can identify in an unsupervised manner have therefore been great interest to simulation community. Here, we developed a general method identifying "on-the-fly" during weighted ensemble (WE) via deep learning (DL) outliers among sampled conformations. Our identifies latent space model system's conformations periodically trained...
Understanding the effects of missense mutations or single amino acid variants (SAVs) on protein function is crucial for elucidating molecular basis diseases/disorders and designing rational therapies. We introduce here Rhapsody-2 , a machine learning tool discriminating pathogenic neutral SAVs, significantly expanding precursor limited by availability structural data. With advent AlphaFold2 as powerful structure prediction, trained expanded dataset 117,525 SAVs corresponding to 12,094 human...
We present the Rate from Event Durations (RED) scheme, a new scheme that more efficiently calculates rate constants using weighted ensemble path sampling strategy. This enables rate-constant estimation shorter trajectories by incorporating probability distribution of event durations, or barrier-crossing times, simulation. have applied RED to simulations variety rare-event processes range in complexity: residue-level protein conformational switching, atomistic Na
We seek to completely revise current models of airborne transmission respiratory viruses by providing never-before-seen atomic-level views the SARS-CoV-2 virus within a aerosol. Our work dramatically extends capabilities multiscale computational microscopy address significant gaps that exist in experimental methods, which are limited their ability interrogate aerosols at atomic/molecular level and thus ob-scure our understanding transmission. demonstrate how integrated data-driven platform...
The catalytic activity of human glutathione S-transferase A1-1 (hGSTA1-1), a homodimeric detoxification enzyme, is dependent on the conformational dynamics key C-terminal helix α9 in each monomer. However, structural details how two monomers interact upon binding substrates not well understood and structure ligand-free state hGSTA1-1 homodimer has been resolved. Here, we used combination electron paramagnetic resonance (EPR) distance measurements weighted ensemble (WE) simulations to...
The weighted ensemble (WE) path sampling strategy has pushed the boundaries of molecular simulation by enabling generation rates and atomistic pathways for biological processes beyond ms timescale. However, WE not yet reached its full potential much can be gained from pursuing stress tests. Here we have explored a test involving seconds-timescale unbinding highly charged ligand protein receptor: release ADP Eg5 receptor, which functions as motor in cell division. From this learned valuable...
We present a new force field, AMBER ff15ipq-m, for simulations of protein mimetics in applications from therapeutics to biomaterials. This field is an expansion the ff15ipq that was developed canonical proteins and enables modeling four classes artificial backbone units are commonly used alongside natural α residues blended or “heterogeneous” backbones: chirality-reversed D-α-residues, Cα-methylated α-residue Aib, homologated β-residues (β3) bearing proteinogenic side chains, two cyclic β...
We introduce WatFinder, a tool designed to identify and visualize protein-water interactions (water bridges, water-mediated associations, or water channels, fluxes, clusters) relevant protein stability, dynamics, function. WatFinder is integrated into ProDy, Python API broadly used for structure-based prediction of dynamics. provides suite functions generating raw data as well outputs from statistical analyses. The ProDy framework facilitates comprehensive automation efficient analysis the...
The weighted ensemble (WE) strategy has been demonstrated to be highly efficient in generating pathways and rate constants for rare events such as protein folding binding using atomistic molecular dynamics simulations. Here we present two sets of tutorials instructing users the best practices preparing, carrying out, analyzing WE simulations various applications WESTPA software. first set more basic describes a range simulation types, from association process explicit solvent complex...
An ongoing challenge to chemists is the analysis of pathways and kinetics for chemical reactions–including transient structures between reactants products that are difficult resolve using laboratory experiments. Here, we enabled direct molecular dynamics simulations a textbook series reactions on hundreds ns µs timescale weighted ensemble (WE) path sampling strategy with hybrid quantum mechanical/molecular mechanical (QM/MM) models. We focused azide-clock involving addition azide anion each...
An ongoing challenge to chemists is the analysis of pathways and kinetics for chemical reactions in solution, including transient structures between reactants products that are difficult resolve using laboratory experiments. Here, we enabled direct molecular dynamics simulations a textbook series on hundreds ns μs time scale weighted ensemble (WE) path sampling strategy with hybrid quantum mechanical/molecular mechanical (QM/MM) models. We focused azide-clock involving addition an azide...