Session Type Key

Hybrid Sessions
Virtual Sessions
In-Person Sessions

custom image


Reset

Advanced Filters

MEDI Poster Session I:
07:00pm - 09:00pm USA / Canada - Pacific - March 20, 2022 | Location: Exhibit Hall C (San Diego Convention Center)
Division: [MEDI] Division of Medicinal Chemistry
Session Type: Poster - In-person
Division/Committee: [MEDI] Division of Medicinal Chemistry
Sunday
3655908 - Synthesizing a-helix mimetics with CADD-positioned electrophilic warheads to inhibit the BFL-1 anti-apoptotic protein
07:00pm - 09:00pm USA / Canada - Pacific - March 20, 2022 | Location: Exhibit Hall C (San Diego Convention Center)
Division: [MEDI] Division of Medicinal Chemistry
Session Type: Poster - In-person
Cells utilize multiple systems to maintain homeostasis. In healthy cells, these systems function normally however, cancer cells can hijack these systems to promote tumor growth. One system that is targeted is the intrinsic apoptosis pathway which is regulated by the B-cell lymphoma 2 (BCL-2) family of proteins. This family is comprised of pro-apoptotic proteins, which include BAK and BAX, and anti-apoptotic proteins, such as BCL-2, BCL-xL, and MCL-1. Inside a healthy cell, these proteins sequester each other preventing cell death. This interaction occurs on the BH3 binding groove found on the surface of anti-apoptotic proteins. Each BH3 groove interacts by recognizing the complementary BH3 “death” domain on pro-apoptotic proteins. A feature of this “death” domain is that it changes into an a-helix structure once bound. This balance is disrupted when a stress is detected, and the cell upregulates pro-apoptotic proteins in response and activation of BAK and BAX causes apoptosis. Cancer cells hijack this system by upregulating anti-apoptotic proteins to such a heightened level that all pro-apoptotic proteins are sequestered, and tumorigenesis occurs. Multiple inhibitors have been designed to target anti-apoptotic proteins via the BH3 binding groove. These have been dubbed “BH3 mimetics”, and they all seek to hoodwink the anti-apoptotic proteins into binding which releases pro-apoptotic proteins causing apoptosis. The only FDA-approved BH3 mimetic is venetoclax, which selectively binds the BCL-2 protein. Additionally, there are multiple clinical trials ongoing for inhibitors that target BCL-xL and MCL-1. Contrarily, BFL-1 is an anti-apoptotic protein that has received less attention, and there are no BFL-1 small-molecule inhibitors. BFL-1 is commonly overexpressed in leukemias, lymphomas, and melanomas while also being recently implicated in venetoclax resistance in leukemias. Unique to BFL-1 is a cysteine residue (Cys55) in its BH3-binding groove, and this residue may be covalently captured in drug design. Here, we will capitalize on previously validated BH3 a-helix mimetic scaffolds and employ computer-aided drug design (CADD) to (i) direct the modification of these scaffolds towards the discovery of selective inhibitors for BFL-1 through non-covalent interactions, as well as (ii) equip our inhibitors with irreversible and reversible covalent electrophilic warheads to react with Cys55. Our efforts towards this goal will be presented.
Sunday
3657306 - Polypharmacologic approach to relapsed/refractory multiple myeloma: Dual inhibition of the proteasome and aggresome pathways
07:00pm - 09:00pm USA / Canada - Pacific - March 20, 2022 | Location: Exhibit Hall C (San Diego Convention Center)
Division: [MEDI] Division of Medicinal Chemistry
Session Type: Poster - In-person
Polypharmacologic agents have increased in popularity, especially in the treatment of multifactorial diseases such as cancer. These drug compounds contain pharmacophores for two or more targets of interest and may result in several advantages over traditional combination treatments, such as a potential synergistic therapeutic affect due to the simultaneous presence of both pharmacophores in the sites of need. The ability of these agents to incapacitate multiple targets can be applied to the treatment of cancer therapeutic resistance, which is often due to a compensatory upregulation of related proteins. Multiple myeloma (MM) is a cancer of the plasma cells, which overproduce abnormal proteins that need to be removed from the cell in order to promote malignant cell survival. For this reason, proteasome inhibitors have become a first-line treatment for MM by inhibiting the main method of protein degradation; however, there is considerable resistance to these treatments, forcing patients into the relapsed/refractory category. One major cause of resistance is the upregulation of histone deacetylase-6 (HDAC6). This enzyme plays a key role in the aggresome pathway, a compensatory protein degradation method. Recently, a phase I/II clinical trial of the combination treatment of bortezomib (proteasome inhibitor) and ricolinostat (HDAC inhibitor) showed therapeutic efficacy in resistant MM. We propose to further these findings by developing dual HDAC6/proteasome inhibitors. Additionally, we are developing compounds with high HDAC6 selectivity; this is hypothesized to decrease the cytotoxic effects of pan-HDAC inhibitors while retaining therapeutic efficacy through inhibition of the aggresome pathway. Furthermore, we are developing novel covalent HDAC6 inhibitors which do not use the traditional hydroxamic acid moiety as a zinc binding group (ZBG). Hydroxamic acids are known to be highly promiscuous binders, contributing to the toxicities seen by HDAC inhibitors. Once developed as an HDAC6 inhibitor, the novel covalent moiety will be applied to our dual inhibitors. We hypothesize that our dual HDAC6/proteasome inhibitors, especially those containing the novel HDAC inhibiting moiety, will result in higher therapeutic effect compared to the traditional combination treatments.
ACS Award for Computers in Chemical & Pharmaceutical Research in honor of Alexander MacKerell (2022) :
08:00am - 12:00pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Alexander Mackerell, Organizer; Henry Woodcock, Organizer; Ashley Ringer, Presider; Jing Huang, Presider
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Division/Committee: [COMP] Division of Computers in Chemistry
Tuesday
3703565 - Efficient utilization of structural information in computer-aided drug design
08:00am - 08:20am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Jing Huang, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Recent advances in computer hardware and software have rendered computer-aided drug design (CADD) more and more powerful. At the same time, there is a huge increase in the number of protein-ligand complex structures available in PDB databank due to advances in structural biology techniques. In this talk, I will present our efforts towards effective and efficient utilization of structural information to enhance the accuracy of CADD within the framework of molecular docking and of molecular generative models, respectively. I will also share some recent results in developing antibacterial and antiviral agents with novel scaffolds.
Tuesday
3703575 - Leveraging protein structure for novel ligand discovery via physics-based approaches OnDemand
08:20am - 08:40am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Dr. Niu Huang, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Molecular docking is now one of the most pragmatic techniques to leverage protein structure for ligand discovery. To strike the balance of efficiency and accuracy, we have systematically improved the accuracy and efficiency of molecular docking by developing a hierarchical strategy to integrate different computational methods in an increasing order of complexity and more physically realistic manner. We here convincingly show that we have arrived at a computational protocol that is effective to discover novel chemotypes conferring new biology. We will demonstrate our applications in different aspects, namely, identifying new lead series for old targets, predicting unknown polypharmacology of old drugs and discovering the first-in-class druggable candidates against novel biological targets. Finally, we will discuss our progress in the evaluation and optimization of our identified novel candidates in different therapeutic areas through a multidisciplinary approach from mechanistic studies to animal models, represents how physics-based protein-ligand binding prediction can serve as the basis of a drug-discovery procedure.
Tuesday
3703651 - Interface of machine learning and drug design
08:40am - 09:00am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
U Deva Priyakumar, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
In this talk, I will present some key developments in machine learning techniques and highlight their novel application in the field of computer-aided drug design.
Tuesday
3703581 - Post translational modifications induced structural reorganization in tau protein
09:00am - 09:20am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Microtubule associated protein Tau (MAPT) is a phospho-protein within neurons of the brain. Aggregation of tau is the leading cause of tauopathies such as Alzheimer’s disease. Tau undergoes several post-translational modifications of which phosphorylation and O-GlcNAcylation are key chemical modifications. Tau aggregates into paired helical filaments (PHF) and neurofibrillary tangles upon hyperphosphorylation whereas O-GlcNAcylation stabilizes the soluble form of Tau. How specific phosphorylation and/or O-GlcNAcylation events influence Tau conformations remains largely unknown due to the disordered nature of Tau. We have investigated the phosphorylation and O-GlcNAcylation induced conformational effects on the proline-rich domain (P2) and the R3-R4 repeat domains of human Tau protein. PTMs were introduced at experimentally verified phosphorylation and O-GlcNAcylation sites. Our studies suggest the opposing structural effects of both PTMs and the importance of salt-bridges in governing the conformational preferences upon phosphorylation, highlighting the role of proximal Arginine and Lysine upon hyper-phosphorylation. The structural effects upon PTM were found to be prominent for phosphorylation when compared to O-GlcNAcylation. One of the key observations from our study was the conformational transition to a more opened “H-conformation” upon phosphorylation when compared to the compact “C-conformation” observed in native Tau PHF. The study sheds light on the structural implication of PTMs paving the way for further investigations.
Tuesday
3703577 - Nucleobase stacking revisited
09:20am - 09:40am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Lennart Nilsson, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
In the mid 1990’s we performed comprehensive studies of the stacking properties of RNA and DNA dinucleotides using PMF calculations with the then current CHARMM nucleic acid all-atom force field.

In the quarter century that has passed the force field has been improved several times, and here a comparison of the performance of the newer CHARMM nucleic acid force fields for nucleobase stacking will be presented.

Tuesday
3703635 - Quantum mechanical methods applied to biological systems
09:40am - 10:00am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Jiali Gao, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
In this talk, I will highlight the history of Quantum Mechanical methods that have been developed and applied to study biological systems with particular emphasis on the intersection of QM methods with QM/MM techniques.
Tuesday
Intermission
10:00am - 10:20am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid

Tuesday
3703628 - Understanding RNA folding
10:20am - 10:40am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Dave Thirumalai, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Folding of RNA involves an interplay between sequence and counterions. I will describe a theoretical approach that accurately describes the thermodynamics and kinetics of ribozyme folding in the presence of magnesium.
Tuesday
3703568 - Conformational constraints versus disorder in liquid-liquid phase separated polypeptide systems
10:40am - 11:00am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Bernard Pettitt, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
We use all atom simulations to consider the role of conformational constraints which change the entropic-enthalpic decomposition in phase separated polypeptides. We compare peptides with and without constraints and compared with experimental solubility studies. We find that constraints can yield surprising changes in the solubility by shifting thermodynamic compensation.
Tuesday
3703631 - Force field and models of biomolecular systems the progress made
11:00am - 11:20am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Benoit Roux, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid

Biological phenomena occur in a variety of environments which can be highly polar and nonpolar. Molecular dynamics (MD) simulations based on atomistic models play an increasingly important role in understanding the fundamental physical forces driving the structure and dynamics of biological membranes. This is essential for the development of drugs for therapeutic purposes. Improvements in these models involve extensions of the underlying functional form of the potential energy, the "force field", as well as additional optimization targeting a wider range of experimental and quantum mechanical data. Membranes present particular challenges because they constitute a molecular environment that varies abruptly over a very short distance, from a high dielectric aqueous region to a low dielectric nonpolar hydrocarbon region. When investigating the physical forces driving molecular processes involving biomembranes, it is important to accurately model the changes in the electronic distribution as changes occur, a need that has stimulated the development of the polarizable models discussed above and being undertaken as part of our research program. Indeed, results show that new physical mechanisms emerge when accounting for induced polarization. The long history of force field development will be reviewed, with a critical eye on the different shortcomings and drawback from different approaches.

Tuesday
3703578 - Ion interactions with DNA G-quadruplexes from Drude polarizable simulations: Hydration, thermodynamics, and competition
11:20am - 11:40am USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
G-quadruplexes (GQs) are noncanonical nucleic acid structures that form in guanine-rich sequences of DNA and RNA. GQs are characterized by stacked guanine tetrads that are stabilized by Hoogsteen hydrogen bonding and coordination of monovalent cations, typically K+. DNA GQs are enriched in promoter regions, origins of replication, and telomeres, suggesting important roles in regulation of gene expression and genome maintenance. As such, GQs may serve as drug targets, particularly against various types of cancer, by modulating GQ folding or interactions with proteins. GQ-specific drug design requires an atomistic understanding of their dynamics, with a particular emphasis on ion interactions, which are crucial for stability. To date, molecular dynamics (MD) simulations have suffered from inaccuracies in representing ion-GQ and ion-ion interactions, due primarily to a lack of induced dipole interactions. We have performed a series of polarizable simulations on structurally diverse DNA GQs, including unbiased MD and metadynamics approaches. Our simulations revealed cooperative polarization of guanine base dipoles in the GQ tetrad core, as well as ion binding to outer faces of GQ tetrads, which impacts accessibility of these sites for small molecules and water exchange. We have also computed the electric field exerted by GQ bases, which directs ions to bind to these locations. Our simulations have also revealed important details about ion competition in the ionic atmosphere around DNA GQs, as well as the extent to which ions must be dehydrated prior to entering the GQ core to become coordinated. The latter simulations reveal critical details about specificity of GQs for K+ over Na+ and Li+. Together, these simulations provide important biophysical insights into the properties of DNA GQs.
Tuesday
3703650 - Multi-scale modeling assessment of drug-induced cardiotoxicity: From atomistic modeling to heart rhythm
11:40am - 12:00pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Igor Vorobyov, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Drug-induced cardiotoxicity in the form of cardiac arrhythmia is a major problem in drug development, which has led to multiple drug candidate attritions from the development or marketplace. It has been attributed to a blockade of cardiac voltage-gated potassium channel Kv11.1, encoded by the human ether-a-go-go-related gene (hERG). However, not all hERG channel blocking drugs cause cardiac arrhythmias, leading to possible abandonment of safe and efficacious pharmaceuticals. We developed multi-scale modeling and simulation pipeline to predict arrhythmogenicity of drugs from their chemical structures. It starts from the development of optimized CHARMM general force field (CGENFF) models of drug molecules in different protonation states and using unbiased and enhanced sampling all-atom molecular dynamics (MD) simulations of their binding to conformational state-specific models of hERG channel structures embedded in hydrated lipid bilayers. MD simulation-derived drug binding affinities as well as their association and dissociation (“on” and “off”) rates were used as functional kinetic model parameters of hERG channel – drug interactions and subsequently used in cardiac cell and tissue simulations to predict emergent drug effects on heart rhythm. This pipeline was successfully used to predict different pro-arrhythmia proclivities of two potent hERG blockers, dofetilide and moxifloxacin. We also elucidated molecular basis of stereospecific arrhythmogenic risks of two enantiomers of beta-blocking anti-arrhythmic drug sotalol. We have been expanding this pipeline to predict how multi-target binding of hERG channel blockers may ameliorate or exaggerate their pro-arrhythmia risks and how sex hormone can modulate drug arrhythmogenicity via direct hERG channel interactions. We are also making this pipeline more robust and high-throughput using Site-Identification by Ligand Competitive Saturation (SILCS) docking and machine learning (ML) techniques. Our in-silico predictive multiscale pipeline will help efficient assessment of cardiac toxicity of diverse drug candidates from their chemical structures.
ACS Award for Computers in Chemical & Pharmaceutical Research in honor of Alexander MacKerell (2022) :
02:00pm - 06:00pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Division/Committee: [COMP] Division of Computers in Chemistry
Tuesday
3703580 - Computing binding affinity at transmembrane region: A case study of PIEZO1 channel agonist
02:00pm - 02:20pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Yun Lyna Luo, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
The development of drugs that act as highly ion channel-specific modulators remains challenging. Unlike soluble proteins, ligands can reach ion channels via hydrophilic pathways (diffuse to the pore from extracellular or intracellular sides) or lipophilic pathways (partition in lipid bilayers first before diffuses to binding sites). While numerous work has been done on the small molecule binding in the aqueous phase, much less attention has been given to ligands that bind to the transmembrane region through the lipophilic pathway. Here we report the first effort in computing a full thermodynamic circle of a small molecule agonist Yoda1 that binds to the transmembrane region of the mechanosensitive channel Piezo1, in a state-dependent manner. We show that hydrophobic molecules that bind to transmembrane regions have higher mobility than traditional protein-ligand interactions anchored by hydrogen bonds. A set of flat-bottom harmonic restraints are designed to capture the high ligand mobility during free energy perturbation while reducing the number of geometric restraints down to two. More importantly, the stronger binding affinity of Yoda1 in Piezo1 open conformation than in closed conformation provided quantitative evidence of the reported Piezo1’s agonist binding site.
Tuesday
3703585 - MDSAPT: Analyzing molecular interactions in molecular dynamics trajectories
02:20pm - 02:40pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Noncovalent interactions play an important role in stabilizing small molecule binding events in macromolecular structures, such as a ligand binding to a protein structure. Often, once the residues involved in stabilizing interactions have been identified, a detailed analysis of the molecular interactions is required, requiring the use of correlated electronic structure methods. To characterize the important interactions over the course of the whole simulation and analyze how these interactions change as the structure changes, snapshots from throughout the simulation would need to be analyzed using the correlated electronic structure methods. This would require a complex workflow, combining trajectory analysis, structure preparation, and quantum chemistry calculations. We have developed MDSAPT, a python package designed to manage the entire workflow of analyzing molecular interactions in a molecular dynamics trajectory, which only requires the user to specify the residues to be analyzed and the pH of the system. The workflow tool can be deployed from the command line or in a jupyter notebook and includes visualization tools.
Tuesday
3703584 - Ring puckering properties of pyranose monosaccharides from human biology
02:40pm - 03:00pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Naturally-occurring glycans in humans are composed of the monosaccharides D-glucose (Glc), N-acetyl-D-glucosamine (GlcNAc), D-galactose (Gal), N-acetyl-D-galactosamine (GalNAc), D-mannose (Man), D-xylose (Xyl), L-fucose (Fuc), N-acetyl-D-neuraminic acid (Neu5Ac), D-glucuronic acid (GlcA), and L-iduronic acid (IdoA), all in their pyranose forms. One determinant of the conformational properties of human glycans are the glycosidic linkages connecting the constituent pyranose monosaccharides; another is the ring puckering of these pyranoses. I will discuss our exhaustive characterization of the ring puckering of both alpha and beta anomers of these monosaccharides along with their respective methyl glycosides using Adaptive Biasing Force molecular dynamics simulations with the CHARMM all-atom fixed-charge carbohydrate force field. Included in the discussion will be an analysis of the accuracy of the force field relative to recent high-quality experimental data, with an emphasis on IdoA and the related molecule, idose, since these two monosaccharides are known to exist in close equilibrium between different puckering conformations. I will also present preliminary data on the same with the CHARMM Drude polarizable force field.
Tuesday
3703582 - Computational approaches to the design and optimization of heterobifunctional degraders
03:00pm - 03:20pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Sean Zhu, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Targeted protein degradation is a modality that selectively directs proteins to the natural protein quality control machinery, the proteosome, for degradation. Although these molecules are larger than the typical small molecules, they have been shown to achieve robust oral exposure and target coverage in humans. Our understanding of the structure-property relationship (SPR) of degrades continues to evolve. Here, we will discuss the challenges and optimization strategies towards addressing potency and ADME/PK properties of degraders. Using an internal program as a case study, we will demonstrate how ternary complex structure predictions help generate design strategies to optimize potency, how computational studies help understand the SPR, and how they can be together leveraged to design orally bioavailable degraders more efficiently.
Tuesday
3703589 - Accurate and efficient relative binding free energy calculations using Lambda dynamics
03:20pm - 03:40pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
E. Prabhu Raman, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Relative free energy methods have the potential to provide accurate estimates of protein-ligand binding affinities given the rigorous physical formulation, and explicit solvent modeling. However, high computational cost limits applications to small sets of compounds. This talk will focus on the lambda dynamics method, which can be substantially more efficient than FEP for screening relatively large combinatorial libraries around the lead compound while maintaining accuracy. An automated workflow will be described that allows set up, execution, and analysis of multi-site lambda dynamics calculations that is implemented in BIOVIA Discovery Studio and Pipeline Pilot, using CHARMM. The workflow establishes a framework for setting up simulation systems for exploratory screening of modifications to a lead compound, enabling the calculation of relative binding affinities of combinatorial libraries. Included in the workflow is an algorithm to automatically optimize biasing potentials that enables enhanced sampling in alchemical space. Retrospective validation of the workflow using a diverse data set of congeneric ligands for seven proteins with experimental binding affinity data will be presented showing the ability of the method to accurately and efficiently predict relative binding affinities.
Tuesday
3649288 - Withdrawn
03:40pm - 04:00pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid

Tuesday
Intermission
04:00pm - 04:20pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid

Tuesday
3703573 - Coarse-grained force field balancing the dynamics of intrinsically disordered peptides and phase behavior in peptide-RNA mixtures
04:20pm - 04:40pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Michael Feig, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Intrinsically disordered peptides (IDPs) are inherently dynamic. They are also often involved in the formation of liquid condensates when present at high concentrations, especially in the presence of co-condensing biomolecules such as nucleic acids. While overall charge and patterning along amino acid chains are main determinants of phase behavior, many details about the driving principles towards condensation as well as the exact molecular nature of condensed environments remain uncertain. Computer simulations are in principle able to investigate these questions, but well-established atomistic models are too costly to reach the spatial scales involved in condensate formation. Coarse-grained (CG) models offer an alternative but require extensive parametrization to be able to carry out predictive simulations. An optimized CG model is presented that originates from generic polymer and colloid models but that was optimized to reproduce recent experiments on phase separation for a variety of peptide/nucleic acid mixtures. The optimized model is used further to determine which sequences lead to the induction of phase separation. Results are presented where this CG model is applied to selected IDPs and phase-separating mixtures.
Tuesday
3703634 - Computer-aided drug discovery using SILCS technology and AI-assisted screening of ultra-large compound library
04:40pm - 05:00pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Sunhwan Jo, Presenter
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Site Identification by Ligand Competitive Saturation (SILCS) is a multiple solute-solvent simulation method for determining functional group affinity patterns for proteins. Using a grand canonical Monte Carlo with molecular dynamics (GCMC-MD) simulation approach, SILCS drives efficient sampling of chemically-diverse probes and water molecules across a protein surface and into deeply buried pockets. The resulting trajectories are processed to create three-dimensional affinity maps (FragMaps) for the different probe atoms. SILCS FragMaps recapitulate functional group patterns of crystallographic ligands across diverse targets, such as nuclear receptors and GPCRs from different gene families, without any target-specific training. FragMap-based Ligand Grid Free Energy (LGFE) is a scoring metric that rapidly rank-orders ligand favorability for any region of a protein. In our recent validation effort with 8 protein targets and 407 ligands with broad chemical variability, it was shown that LGFE scoring can predict the experimental free energy change in 77% of the times. Recent development of machine learning algorithms and deep learning algorithms can further enhance the utility of SILCS based technology in CADD. For example, the LGFE scoring can be further improved by machine learning algorithms, which can improve the accuracy to 82% in the same benchmark. In addition, a deep learning algorithm is being developed to predict the FragMap, which could greatly improves the availability of FragMap without a lengthy simulation. Finally, active learning based virtual screening will be discussed, which could tackle screening huge library of chemical database greater than 109 compounds.
Tuesday
3703586 - Accurate and fast predictions of reactive cysteine and lysine locations for targeted covalent drug design
05:00pm - 05:20pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
In recent years, targeted covalent inhibitor design has gained widespread interest in drug discovery programs as a way to improve selectivity and potency. In such a design, an electrophilic warhead is added to a reversible inhibitor to covalently engage a nucleophilic residue in the target protein. Thus, it is desirable to develop a computational tool to predict and rationalize highly nucleophilic sites. In this talk, I will discuss our recent efforts in developing and applying the GPU-accelerated continuous constant pH molecular dynamics tool to retro- and prospectively predict reactive cysteine and lysine locations in a large number of kinases, including 14 MAP kinases and all kinases that possess a so-called Ncap cysteine. Time permitted, i will also present our most result findings of reactive residues in KRAS proteins.
Tuesday
3652606 - From CHARMM to SILCS: Confluence of force fields and computer aided drug design
05:20pm - 05:55pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid
Development of empirical force fields including the additive all-atom CHARMM force field and more recently the polarizable classical Drude oscillator model has consumed a significant part of my life over that last 35 years. Performing force field optimization necessarily exposes one to subtle aspects of the physical forces dictating a wide range of chemical and biological phenomena, including intermolecular interactions important in the area of drug design. Awareness of the details of physical forces driving intermolecular interactions facilitated the discovery and development of the SILCS (site-identification by ligand competitive saturation) approach. SILCS represents an alternative to the majority of ligand-binding computational methods by generating a pre-computed ensemble of functional group interactions obtained through molecular simulations of solute-protein interactions in the presence of an explicit solvent representation. Such an approach takes advantage of the high accuracy of the force field parameters in treating the balance between solute-water, solute-protein and water-protein interactions combined with the ability to extensively sample the ensemble of the distribution of the solutes and water around the protein as well as of the conformations of the protein. The resulting functional group affinity pattern, or FragMaps, may then be used for a range of aspects of drug design from database screening through ligand optimization to the formulation of biologics. Interestingly, considering our efforts in force field parametrization, the SILCS pre-computed ensemble strategy greatly diminishes the sensitivity of predictions of drug-like ligand affinity to details of the underlying force field. Our efforts in force field development and drug design have been facilitated by the support and interactions with a large collection of colleagues; various aspects of those efforts will be presented.
Tuesday
Concluding Remarks
05:55pm - 06:00pm USA / Canada - Pacific - March 22, 2022 | Location: Room 25B (San Diego Convention Center)
Division: [COMP] Division of Computers in Chemistry
Session Type: Oral - Hybrid