DNA binding analysis of rare variants in homeodomains reveals homeodomain specificity-determining residues.

Homeodomains (HDs) are the second largest class of DNA binding domains (DBDs) among eukaryotic sequence-specific transcription factors (TFs) and are the TF structural class with the largest number of disease-associated mutations in the Human Gene Mutation Database (HGMD). Despite numerous structural studies and large-scale analyses of HD DNA binding specificity, HD-DNA recognition is still not fully understood. Here, we analyze 92 human HD mutants, including disease-associated variants and variants of uncertain significance (VUS), for their effects on DNA binding activity. Many of the variants alter DNA binding affinity and/or specificity. Detailed biochemical analysis and structural modeling identifies 14 previously unknown specificity-determining positions, 5 of which do not contact DNA. The same missense substitution at analogous positions within different HDs often exhibits different effects on DNA binding activity. Variant effect prediction tools perform moderately well in distinguishing variants with altered DNA binding affinity, but poorly in identifying those with altered binding specificity. Our results highlight the need for biochemical assays of TF coding variants and prioritize dozens of variants for further investigations into their pathogenicity and the development of clinical diagnostics and precision therapies.

DFT/TDDFT calculations of geometry optimization, electronic structure and spectral properties of clevudine and telbivudine for treatment of chronic hepatitis B.

Chronic hepatitis B remains a worldwide health concern. Presently, many drugs, such as Clevudine and Telbivudine, are recommended for the treatment of chronic hepatitis B disease. For this purpose, the quantum chemical analysis of ELUMO-HOMO (Egap), ionization potential (IP), electron affinity (EA), electronegativity (EN), chemical hardness (η), chemical potential (µ), chemical softness (S), electrophilicity index (ω), electron accepting capability (ω+), electron-donating capability (ω-), Nucleophilicity index (N), additional electronic charge (∆Nmax), Optical softness (σ0) and Dipole Moment, IR and UV-Vis spectra, molecular electrostatic potential (MEP) profile, Mulliken charge analysis, natural bond orbital (NBO) were examined in this study. The dipole moment of the compounds suggests their binding pose and predicted binding affinity. The electrophilic and nucleophilic regions were identified, and techniques such as NBO, UV-Vis, and IR were used to gain insights into the molecular structure, electronic transitions, and potential drug design for Hepatitis B treatment. Calculations for this study were carried out using the Gaussian 09 program package coupled with the DFT/TDDFT technique. The hybrid B3LYP functional method and the 6-311++G(d, p) basis set were used for the calculations.

The molecular basis of Abelson kinase regulation by its αI-helix.

Abelson tyrosine kinase (Abl) is regulated by the arrangement of its regulatory core, consisting sequentially of the SH3, SH2, and kinase (KD) domains, where an assembled or disassembled core corresponds to low or high kinase activity, respectively. It was recently established that binding of type II ATP site inhibitors, such as imatinib, generates a force from the KD N-lobe onto the SH3 domain and in consequence disassembles the core. Here, we demonstrate that the C-terminal αI-helix exerts an additional force toward the SH2 domain, which correlates both with kinase activity and type II inhibitor-induced disassembly. The αI-helix mutation E528K, which is responsible for the ABL1 malformation syndrome, strongly activates Abl by breaking a salt bridge with the KD C-lobe and thereby increasing the force onto the SH2 domain. In contrast, the allosteric inhibitor asciminib strongly reduces Abl's activity by fixating the αI-helix and reducing the force onto the SH2 domain. These observations are explained by a simple mechanical model of Abl activation involving forces from the KD N-lobe and the αI-helix onto the KD/SH2SH3 interface.

Transient interdomain interactions in free USP14 shape its conformational ensemble.

The deubiquitinase (DUB) ubiquitin-specific protease 14 (USP14) is a dual domain protein that plays a regulatory role in proteasomal degradation and has been identified as a promising therapeutic target. USP14 comprises a conserved USP domain and a ubiquitin-like (Ubl) domain separated by a 25-residue linker. The enzyme activity of USP14 is autoinhibited in solution, but is enhanced when bound to the proteasome, where the Ubl and USP domains of USP14 bind to the Rpn1 and Rpt1/Rpt2 units, respectively. No structure of full-length USP14 in the absence of proteasome has yet been presented, however, earlier work has described how transient interactions between Ubl and USP domains in USP4 and USP7 regulate DUB activity. To better understand the roles of the Ubl and USP domains in USP14, we studied the Ubl domain alone and in full-length USP14 by nuclear magnetic resonance spectroscopy and used small angle x-ray scattering and molecular modeling to visualize the entire USP14 protein ensemble. Jointly, our results show how transient interdomain interactions between the Ubl and USP domains of USP14 predispose its conformational ensemble for proteasome binding, which may have functional implications for proteasome regulation and may be exploited in the design of future USP14 inhibitors.

Synthesis, crystal structure and in-silico evaluation of arylsulfonamide Schiff bases for potential activity against colon cancer.

This report presents a comprehensive investigation into the synthesis and characterization of Schiff base compounds derived from benzenesulfonamide. The synthesis process, involved the reaction between N-cycloamino-2-sulfanilamide and various substituted o-salicylaldehydes, resulted in a set of compounds that were subjected to rigorous characterization using advanced spectral techniques, including 1H NMR, 13C NMR and FT-IR spectroscopy, and single-crystal X-ray diffraction. Furthermore, an in-depth assessment of the synthesized compounds was conducted through Absorption, Distribution, Metabolism, Excretion and Toxicity (ADMET) analysis, in conjunction with docking studies, to elucidate their pharmacokinetic profiles and potential. Impressively, the ADMET analysis showcased encouraging drug-likeness properties of the newly synthesized Schiff bases. These computational findings were substantiated by molecular properties derived from density functional theory (DFT) calculations using the B3LYP/6-31G* method within the Jaguar Module of Schrödinger 2023-2 from Maestro (Schrodinger LLC, New York, USA). The exploration of frontier molecular orbitals (HOMO and LUMO) enabled the computation of global reactivity descriptors (GRDs), encompassing charge separation (Egap) and global softness (S). Notably, within this analysis, one Schiff base, namely, 4-bromo-2-{N-[2-(pyrrolidine-1-sulfonyl)phenyl]carboximidoyl}phenol, 20, emerged with the smallest charge separation (ΔEgap = 3.5780 eV), signifying heightened potential for biological properties. Conversely, 4-bromo-2-{N-[2-(piperidine-1-sulfonyl)phenyl]carboximidoyl}phenol, 17, exhibited the largest charge separation (ΔEgap = 4.9242 eV), implying a relatively lower propensity for biological activity. Moreover, the synthesized Schiff bases displayed remarkeable inhibition of tankyrase poly(ADP-ribose) polymerase enzymes, integral in colon cancer, surpassing the efficacy of a standard drug used for the same purpose. Additionally, their bioavailability scores aligned closely with established medications such as trifluridine and 5-fluorouracil. The exploration of molecular electrostatic potential through colour mapping delved into the electronic behaviour and reactivity tendencies intrinsic to this diverse range of molecules.

Using atomistic solution scattering modelling to elucidate the role of the Fc glycans in human IgG4.

Human immunoglobulin G (IgG) exists as four subclasses IgG1-4, each of which has two Fab subunits joined by two hinges to a Fc subunit. IgG4 has the shortest hinge with 12 residues. The Fc subunit has two glycan chains, but the importance of glycosylation is not fully understood in IgG4. Here, to evaluate the stability and structure of non-glycosylated IgG4, we performed a multidisciplinary structural study of glycosylated and deglycosylated human IgG4 A33 for comparison with our similar study of human IgG1 A33. After deglycosylation, IgG4 was found to be monomeric by analytical ultracentrifugation; its sedimentation coefficient of 6.52 S was reduced by 0.27 S in reflection of its lower mass. X-ray and neutron solution scattering showed that the overall Guinier radius of gyration RG and its cross-sectional values after deglycosylation were almost unchanged. In the P(r) distance distribution curves, the two M1 and M2 peaks that monitor the two most common distances within IgG4 were unchanged following deglycosylation. Further insight from Monte Carlo simulations for glycosylated and deglycosylated IgG4 came from 111,382 and 117,135 possible structures respectively. Their comparison to the X-ray and neutron scattering curves identified several hundred best-fit models for both forms of IgG4. Principal component analyses showed that glycosylated and deglycosylated IgG4 exhibited different conformations from each other. Within the constraint of unchanged RG and M1-M2 values, the glycosylated IgG4 models showed more restricted Fc conformations compared to deglycosylated IgG4, but no other changes. Kratky plots supported this interpretation of greater disorder upon deglycosylation, also observed in IgG1. Overall, these more variable Fc conformations may demonstrate a generalisable impact of deglycosylation on Fc structures, but with no large conformational changes in IgG4 unlike those seen in IgG1.

Do antibody CDR loops change conformation upon binding?

Antibodies have increasingly been developed as drugs with over 100 now licensed in the US or EU. During development, it is often necessary to increase or reduce the affinity of an antibody and rational attempts to do so rely on having a structure of the antibody-antigen complex often obtained by modeling. The antigen-binding site consists primarily of six loops known as complementarity-determining regions (CDRs), and an open question has been whether these loops change their conformation when they bind to an antigen. Existing surveys of antibody-antigen complex structures have only examined CDR conformational change in case studies or small-scale surveys. With an increasing number of antibodies where both free and complexed structures have been deposited in the Protein Data Bank, a large-scale survey of CDR conformational change during binding is now possible. To this end, we built a dataset, AbAgDb, that currently includes 177 antibodies with high-quality CDRs, each of which has at least one bound and one unbound structure. We analyzed the conformational change of the Cα backbone of each CDR upon binding and found that, in most cases, the CDRs (other than CDR-H3) show minimal movement, while 70.6% and 87% of CDR-H3s showed global Cα RMSD ≤ 1.0Å and ≤ 2.0Å, respectively. We also compared bound CDR conformations with the conformational space of unbound CDRs and found most of the bound conformations are included in the unbound conformational space. In future, our results will contribute to developing insights into antibodies and new methods for modeling and docking.

The structural landscape of the immunoglobulin fold by large-scale de novo design.

De novo designing immunoglobulin-like frameworks that allow for functional loop diversification shows great potential for crafting antibody-like scaffolds with fully customizable structures and functions. In this work, we combined de novo parametric design with deep-learning methods for protein structure prediction and design to explore the structural landscape of 7-stranded immunoglobulin domains. After screening folding of nearly 4 million designs, we have assembled a structurally diverse library of ~50,000 immunoglobulin domains with high-confidence AlphaFold2 predictions and structures diverging from naturally occurring ones. The designed dataset enabled us to identify structural requirements for the correct folding of immunoglobulin domains, shed light on ß-sheet-ß-sheet rotational preferences and how these are linked to functional properties. Our approach eliminates the need for preset loop conformations and opens the route to large-scale de novo design of immunoglobulin-like frameworks.

In Vivo Disulfide-Bond Crosslinking to Study ß-Barrel Membrane Protein Interactions, Dynamicity, and Folding Intermediates.

Membrane-embedded ß-barrels are the major building blocks of the Gram-negative outer membrane and are involved in antibiotic resistance, virulence, and the maintenance of bacterial cell physiology. The increased frequency of multidrug resistant Gram-negative infections warrants the sharing of accessible methods for the study of ß-barrels. One such method is "in vivo disulfide-bond crosslinking" which is a highly informative and cost-effective approach to study the structure, topology, dynamicity, and function of ß-barrels in situ. The approach can also be used to identify and finely map both stable or transient interactions between ß-barrels and other interacting proteins. In this chapter, I describe the conceptual basis of in vivo disulfide-bond crosslinking and the potential pitfalls in experimental design. I also provide a general protocol for high-efficiency in vivo disulfide-bond crosslinking and modified protocols as examples for how the method can be adapted to different scenarios.

Investigating the anticancer potential of 4-phenylthiazole derived Ru(II) and Os(II) metalacycles.

In this contribution we report the synthesis, characterization and in vitro anticancer activity of novel cyclometalated 4-phenylthiazole-derived ruthenium(II) (2a-e) and osmium(II) (3a-e) complexes. Formation and sufficient purity of the complexes were unambigiously confirmed by 1H-, 13C- and 2D-NMR techniques, X-ray diffractometry, HRMS and elemental analysis. The binding preferences of these cyclometalates to selected amino acids and to DNA models including G-quadruplex structures were analyzed. Additionally, their stability and behaviour in aqueous solutions was determined by UV-Vis spectroscopy. Their cellular accumulation, their ability of inducing apoptosis, as well as their interference in the cell cycle were studied in SW480 colon cancer cells. The anticancer potencies were investigated in three human cancer cell lines and revealed IC50 values in the low micromolar range, in contrast to the biologically inactive ligands.

Synthesis of a 3,7-Disubstituted Isothiazolo[4,3-b]pyridine as a Potential Inhibitor of Cyclin G-Associated Kinase.

Disubstituted isothiazolo[4,3-b]pyridines are known inhibitors of cyclin G-associated kinase. Since 3-substituted-7-aryl-isothiazolo[4,3-b]pyridines remain elusive, a strategy was established to prepare this chemotype, starting from 2,4-dichloro-3-nitropyridine. Selective C-4 arylation using ligand-free Suzuki-Miyaura coupling and palladium-catalyzed aminocarbonylation functioned as key steps in the synthesis. The 3-N-morpholinyl-7-(3,4-dimethoxyphenyl)-isothiazolo[4,3-b]pyridine was completely devoid of GAK affinity, in contrast to its 3,5- and 3,6-disubstituted congeners. Molecular modeling was applied to rationalize its inactivity as a GAK ligand.

Structure-guided engineering and molecular simulations to design a potent monoclonal antibody to target aP2 antigen for adaptive immune response instigation against type 2 diabetes.

Introduction: Diabetes mellitus (DM) is recognized as one of the oldest chronic diseases and has become a significant public health issue, necessitating innovative therapeutic strategies to enhance patient outcomes. Traditional treatments have provided limited success, highlighting the need for novel approaches in managing this complex disease. Methods: In our study, we employed graph signature-based methodologies in conjunction with molecular simulation and free energy calculations. The objective was to engineer the CA33 monoclonal antibody for effective targeting of the aP2 antigen, aiming to elicit a potent immune response. This approach involved screening a mutational landscape comprising 57 mutants to identify modifications that yield significant enhancements in binding efficacy and stability. Results: Analysis of the mutational landscape revealed that only five substitutions resulted in noteworthy improvements. Among these, mutations T94M, A96E, A96Q, and T94W were identified through molecular docking experiments to exhibit higher docking scores compared to the wild-type. Further validation was provided by calculating the dissociation constant (KD), which showed a similar trend in favor of these mutations. Molecular simulation analyses highlighted T94M as the most stable complex, with reduced internal fluctuations upon binding. Principal components analysis (PCA) indicated that both the wild-type and T94M mutant displayed similar patterns of constrained and restricted motion across principal components. The free energy landscape analysis underscored a single metastable state for all complexes, indicating limited structural variability and potential for high therapeutic efficacy against aP2. Total binding free energy (TBE) calculations further supported the superior performance of the T94M mutation, with TBE values demonstrating the enhanced binding affinity of selected mutants over the wild-type. Discussion: Our findings suggest that the T94M substitution, along with other identified mutations, significantly enhances the therapeutic potential of the CA33 antibody against DM by improving its binding affinity and stability. These results not only contribute to a deeper understanding of antibody-antigen interactions in the context of DM but also provide a valuable framework for the rational design of antibodies aimed at targeting this disease more effectively.

Investigating the ability of deep learning-based structure prediction to extrapolate and/or enrich the set of antibody CDR canonical forms.

Deep learning models have been shown to accurately predict protein structure from sequence, allowing researchers to explore protein space from the structural viewpoint. In this paper we explore whether "novel" features, such as distinct loop conformations can arise from these predictions despite not being present in the training data. Here we have used ABodyBuilder2, a deep learning antibody structure predictor, to predict the structures of ~1.5M paired antibody sequences. We examined the predicted structures of the canonical CDR loops and found that most of these predictions fall into the already described CDR canonical form structural space. We also found a small number of "new" canonical clusters composed of heterogeneous sequences united by a common sequence motif and loop conformation. Analysis of these novel clusters showed their origins to be either shapes seen in the training data at very low frequency or shapes seen at high frequency but at a shorter sequence length. To evaluate explicitly the ability of ABodyBuilder2 to extrapolate, we retrained several models whilst withholding all antibody structures of a specific CDR loop length or canonical form. These "starved" models showed evidence of generalisation across CDRs of different lengths, but they did not extrapolate to loop conformations which were highly distinct from those present in the training data. However, the models were able to accurately predict a canonical form even if only a very small number of examples of that shape were in the training data. Our results suggest that deep learning protein structure prediction methods are unable to make completely out-of-domain predictions for CDR loops. However, in our analysis we also found that even minimal amounts of data of a structural shape allow the method to recover its original predictive abilities. We have made the ~1.5 M predicted structures used in this study available to download at https://doi.org/10.5281/zenodo.10280181.

Synthesis, characterization, molecular modeling, binding energies of ß-cyclodextrin-inclusion complexes of quercetin: Modification of photo physical behavior upon ß-CD complexation.

We prepared a naturally occurring flavanoid namely quercetin from tea leaves and analyzed by Absorption, Emission, FT-IR, 1H, 13C nmr spectra and ESI-MS analysis. The inclusion behavior of quercetin in cyclodextrins like α-, ß-, γ-, per-6-ABCD and mono-6-ABCD cavities were supported such as UV-vis., Emission, FT-IR and ICD spectra and energy minimization studies. From the absorption and emission results, the type of complexes formed were found to depend on stoichiometry of Host:Guest. FT-IR data of CD complexes of quercetin supported inclusion complex formation of the substrate with α-, ß- and γ-CDs. The inclusion of host-guest complexation of quercetin with α-, ß-, γ-CDs, per-6-ABCD and mono-6-ABCDs provides very valuable information about the CD:quercetin complexes, the study also shows that ß-CD complexation improves water solubility, chemical stability and bioavailability of quercetin. Besides, phase solubility studies also supported the formation of 1:1 drug-CD soluble complexes. All these spectral results provide insight into the binding behavior of substrate into CD cavity in the order per-6-ABCD > Mono-6-ABCD > γ-CD > ß-CD > α-CD. The proposed model also finds strong support from the fact with excess CD this exciton coupling disappears indicates the formation of only 1:1 complex.

Rapid and automated design of two-component protein nanomaterials using ProteinMPNN.

The design of protein-protein interfaces using physics-based design methods such as Rosetta requires substantial computational resources and manual refinement by expert structural biologists. Deep learning methods promise to simplify protein-protein interface design and enable its application to a wide variety of problems by researchers from various scientific disciplines. Here, we test the ability of a deep learning method for protein sequence design, ProteinMPNN, to design two-component tetrahedral protein nanomaterials and benchmark its performance against Rosetta. ProteinMPNN had a similar success rate to Rosetta, yielding 13 new experimentally confirmed assemblies, but required orders of magnitude less computation and no manual refinement. The interfaces designed by ProteinMPNN were substantially more polar than those designed by Rosetta, which facilitated in vitro assembly of the designed nanomaterials from independently purified components. Crystal structures of several of the assemblies confirmed the accuracy of the design method at high resolution. Our results showcase the potential of deep learning-based methods to unlock the widespread application of designed protein-protein interfaces and self-assembling protein nanomaterials in biotechnology.

MolFeSCue: enhancing molecular property prediction in data-limited and imbalanced contexts using few-shot and contrastive learning.

MOTIVATION: Predicting molecular properties is a pivotal task in various scientific domains, including drug discovery, material science, and computational chemistry. This problem is often hindered by the lack of annotated data and imbalanced class distributions, which pose significant challenges in developing accurate and robust predictive models. RESULTS: This study tackles these issues by employing pretrained molecular models within a few-shot learning framework. A novel dynamic contrastive loss function is utilized to further improve model performance in the situation of class imbalance. The proposed MolFeSCue framework not only facilitates rapid generalization from minimal samples, but also employs a contrastive loss function to extract meaningful molecular representations from imbalanced datasets. Extensive evaluations and comparisons of MolFeSCue and state-of-the-art algorithms have been conducted on multiple benchmark datasets, and the experimental data demonstrate our algorithm's effectiveness in molecular representations and its broad applicability across various pretrained models. Our findings underscore MolFeSCues potential to accelerate advancements in drug discovery. AVAILABILITY AND IMPLEMENTATION: We have made all the source code utilized in this study publicly accessible via GitHub at http://www.healthinformaticslab.org/supp/ or https://github.com/zhangruochi/MolFeSCue. The code (MolFeSCue-v1-00) is also available as the supplementary file of this paper.

Diverging co-translational protein complex assembly pathways are governed by interface energy distribution.

Protein-protein interactions are at the heart of all cellular processes, with the ribosome emerging as a platform, orchestrating the nascent-chain interplay dynamics. Here, to study the characteristics governing co-translational protein folding and complex assembly, we combine selective ribosome profiling, imaging, and N-terminomics with all-atoms molecular dynamics. Focusing on conserved N-terminal acetyltransferases (NATs), we uncover diverging co-translational assembly pathways, where highly homologous subunits serve opposite functions. We find that only a few residues serve as "hotspots," initiating co-translational assembly interactions upon exposure at the ribosome exit tunnel. These hotspots are characterized by high binding energy, anchoring the entire interface assembly. Alpha-helices harboring hotspots are highly thermolabile, folding and unfolding during simulations, depending on their partner subunit to avoid misfolding. In vivo hotspot mutations disrupted co-translational complexation, leading to aggregation. Accordingly, conservation analysis reveals that missense NATs variants, causing neurodevelopmental and neurodegenerative diseases, disrupt putative hotspot clusters. Expanding our study to include phosphofructokinase, anthranilate synthase, and nucleoporin subcomplex, we employ AlphaFold-Multimer to model the complexes' complete structures. Computing MD-derived interface energy profiles, we find similar trends. Here, we propose a model based on the distribution of interface energy as a strong predictor of co-translational assembly.

Mechanistic Insights Revealed by YbtPQ in the Occluded State.

Recently, several ATP-binding cassette (ABC) importers have been found to adopt the typical fold of type IV ABC exporters. Presumably, these importers would function under the transport scheme of "alternating access" like those exporters, cycling through inward-open, occluded, and outward-open conformations. Understanding how the exporter-like importers move substrates in the opposite direction requires structural studies on all the major conformations. To shed light on this, here we report the structure of yersiniabactin importer YbtPQ from uropathogenic Escherichia coli in the occluded conformation trapped by ADP-vanadate (ADP-Vi) at a 3.1 Å resolution determined by cryo-electron microscopy. The structure shows unusual local rearrangements in multiple helices and loops in its transmembrane domains (TMDs). In addition, the dimerization of the nucleotide-binding domains (NBDs) promoted by the vanadate trapping is highlighted by the "screwdriver" action at one of the two hinge points. These structural observations are rare and thus provide valuable information to understand the structural plasticity of the exporter-like ABC importers.

Molecular Basis for the Involvement of Mammalian Serum Albumin in the AGE/RAGE Axis: A Comprehensive Computational Study.

In mammals, glycated serum albumin (gSA) contributes to the pathogenesis of many metabolic diseases by activating the receptors (RAGE) for advanced glycation end products (AGEs). Many aspects of the gSA-RAGE interaction remain unknown. The purpose of the present paper was to study the interaction of glycated human albumin (gHSA) with RAGE using molecular modeling methods. Ten models of gHSA modified with different lysine residues to carboxymethyl-lysines were prepared. Complexes of gHSA-RAGE were obtained by the macromolecular docking method with subsequent molecular dynamics simulation (MD). According to the MD, the RAGE complexes with gHSA glycated at Lys233, Lys64, Lys525, Lys262 and Lys378 are the strongest. Three-dimensional models of the RAGE dimers with gHSA were proposed. Additional computational experiments showed that the binding of fatty acids (FAs) to HSA does not affect the ability of Lys525 (the most reactive lysine) to be glycated. In contrast, modification of Lys525 reduces the affinity of albumin for FA. The interspecies differences in the molecular structure of albumin that may affect the mechanism of the gSA-RAGE interaction were discussed. The obtained results will help us to learn more about the molecular basis for the involvement of serum albumin in the AGE/RAGE axis and improve the methodology for studying cellular signaling pathways involving RAGE.

Analysis of the Conformational Landscape of the N-Domains of the AAA ATPase p97: Disentangling the Continuous Conformational Variability in Partially Symmetrical Complexes.

Single-particle cryo-electron microscopy (cryo-EM) has been shown to be effective in defining the structure of macromolecules, including protein complexes. Complexes adopt different conformations and compositions to perform their biological functions. In cryo-EM, the protein complexes are observed in solution, enabling the recording of images of the protein in multiple conformations. Various methods exist for capturing the conformational variability through analysis of cryo-EM data. Here, we analyzed the conformational variability in the hexameric AAA + ATPase p97, a complex with a six-fold rotational symmetric core surrounded by six flexible N-domains. We compared the performance of discrete classification methods with our recently developed method, MDSPACE, which uses 3D-to-2D flexible fitting of an atomic structure to images based on molecular dynamics (MD) simulations. Our analysis detected a novel conformation adopted by approximately 2% of the particles in the dataset and determined that the N-domains of p97 sway by up to 60° around a central position. This study demonstrates the application of MDSPACE in analyzing the continuous conformational changes in partially symmetrical protein complexes, systems notoriously difficult to analyze due to the alignment errors caused by their partial symmetry.