Interchain Contacts Research Articles

Kinetics of Polyampholyte Dimerization: Influence of Charge Sequences

Polyampholytes (PAs) exhibit complex behaviors in various environments influenced by their charge distribution. This study focuses on the kinetics of dimerization of PAs, aiming to elucidate the underlying mechanisms and clarify relevant characteristics of the charge sequence. We focus on PAs with non-zero net charges, employing molecular dynamics simulations and theoretical analyses to examine how charge sequences influence the rates of dimer formation and dissociation. Our findings reveal that the charge sequence of tails and the blockiness of the minority charge group markedly influence the kinetics of dimerization: large blockiness and tails with a high number of majority-type charges slow down the dissociation of dimers. Additionally, the presence of an extended (central) block of the majority charge promotes structural diversity. Within dimer states, blocks alternate between intra- and inter-chain contacts. The duration times in the dimer states are significantly longer than the typical dwell times of block inter-contacts, with a notable extension when multiple blocks are engaged. Intrinsically disordered proteins (IDPs) play crucial roles in cellular functions, primarily due to their ability to undergo rapid conformational changes and form transient complexes. These properties largely depend on the sequence of charged residues. We provide insights into the fundamental principles governing the structural and dynamical properties of polyampholytic IDP, emphasizing the importance of sequence-specific effects on both aggregation and dissociation.

Direct computations of viscoelastic moduli of biomolecular condensates.

Biomolecular condensates are viscoelastic materials defined by time-dependent, sequence-specific complex shear moduli. Here, we show that viscoelastic moduli can be computed directly using a generalization of the Rouse model that leverages information regarding intra- and inter-chain contacts, which we extract from equilibrium configurations of lattice-based Metropolis Monte Carlo (MMC) simulations of phase separation. The key ingredient of the generalized Rouse model is a graph Laplacian that we compute from equilibrium MMC simulations. We compute two flavors of graph Laplacians, one based on a single-chain graph that accounts only for intra-chain contacts, and the other referred to as a collective graph that accounts for inter-chain interactions. Calculations based on the single-chain graph systematically overestimate the storage and loss moduli, whereas calculations based on the collective graph reproduce the measured moduli with greater fidelity. However, in the long time, low-frequency domain, a mixture of the two graphs proves to be most accurate. In line with the theory of Rouse and contrary to recent assertions, we find that a continuous distribution of relaxation times exists in condensates. The single crossover frequency between dominantly elastic vs dominantly viscous behaviors does not imply a single relaxation time. Instead, it is influenced by the totality of the relaxation modes. Hence, our analysis affirms that viscoelastic fluid-like condensates are best described as generalized Maxwell fluids. Finally, we show that the complex shear moduli can be used to solve an inverse problem to obtain the relaxation time spectra that underlie the dynamics within condensates. This is of practical importance given advancements in passive and active microrheology measurements of condensate viscoelasticity.

PCP-GC-LM: single-sequence-based protein contact prediction using dual graph convolutional neural network and convolutional neural network

BackgroundRecently, the process of evolution information and the deep learning network has promoted the improvement of protein contact prediction methods. Nevertheless, still remain some bottleneck: (1) One of the bottlenecks is the prediction of orphans and other fewer evolution information proteins. (2) The other bottleneck is the method of predicting single-sequence-based proteins mainly focuses on selecting protein sequence features and tuning the neural network architecture, However, while the deeper neural networks improve prediction accuracy, there is still the problem of increasing the computational burden. Compared with other neural networks in the field of protein prediction, the graph neural network has the following advantages: due to the advantage of revealing the topology structure via graph neural network and being able to take advantage of the hierarchical structure and local connectivity of graph neural networks has certain advantages in capturing the features of different levels of abstraction in protein molecules. When using protein sequence and structure information for joint training, the dependencies between the two kinds of information can be better captured. And it can process protein molecular structures of different lengths and shapes, while traditional neural networks need to convert proteins into fixed-size vectors or matrices for processing.ResultsHere, we propose a single-sequence-based protein contact map predictor PCP-GC-LM, with dual-level graph neural networks and convolution networks. Our method performs better with other single-sequence-based predictors in different independent tests. In addition, to verify the validity of our method against complex protein structures, we will also compare it with other methods in two homodimers protein test sets (DeepHomo test dataset and CASP-CAPRI target dataset). Furthermore, we also perform ablation experiments to demonstrate the necessity of a dual graph network. In all, our framework presents new modules to accurately predict inter-chain contact maps in protein and it’s also useful to analyze interactions in other types of protein complexes.

Deciphering the impact of F23L mutation on the aggregation propensity of human islet amyloid polypeptide using molecular simulations

The self-aggregation of 37-residue human islet amyloid polypeptide (hIAPP) into β-sheet rich oligomers and amyloid fibrils contributes to the loss of pancreatic β-cell mass in type 2 diabetes (T2D). A previous study highlighted that the F23L mutation in the amyloidogenic core (Ser20–Ser29) of hIAPP slows down its self-aggregation by a factor of 2.6. However, the underlying molecular mechanism behind slower aggregation and reduced cytotoxicity of hIAPP on F23L mutation remains obscure. Thus, the structural and conformational changes in F23L–hIAPP that lead to slower aggregation and lower cytotoxicity were examined by molecular dynamics (MD) simulations. The MD simulations depicted that F23L mutation marginally increases the helical content from 10.20 ± 1.94 to 11.50 ± 2.18 % accompanied by a noteworthy decrease in aggregation-prone β-sheet from 14.00 ± 1.58 to 0.80 ± 0.37 %. Notably, FEL conformations depicted a complete disappearance of β-sheet content in F23L–hIAPP as compared to 9–20 % in wt–hIAPP in agreement with the notably reduced toxicity of F23L–hIAPP aggregates against cultured INS-1 832/13 cells. Interestingly, dihedral principal component analysis (dPCA) depicted noteworthy compression in the conformational sub-space and a lower trace value of hIAPP on F23L mutation, which indicates lower conformational fluctuations in F23L–hIAPP. F23L mutation alters the conformational landscape of hIAPP dimer by lowering interchain contacts, and hydrogen bonds as well as reducing binding affinity between chains A-B of hIAPP dimer. Importantly, MD simulations illuminated the key role of Phe23 in mediating intramolecular interactions leading to aggregation-prone structures that promote hIAPP aggregation. The key insights regarding structural transitions and conformational preferences of hIAPP on F23L mutation illuminated from this work will benefit the design and optimization of potential inhibitors targeting Phe23-mediated interactions to block hIAPP aggregation in T2D.

Differential Effects of Sequence-Local versus Nonlocal Charge Patterns on Phase Separation and Conformational Dimensions of Polyampholytes as Model Intrinsically Disordered Proteins.

Conformational properties of intrinsically disordered proteins (IDPs) are governed by a sequence-ensemble relationship. To differentiate the impact of sequence-local versus sequence-nonlocal features of an IDP's charge pattern on its conformational dimensions and its phase-separation propensity, the charge "blockiness" κ and the nonlocality-weighted sequence charge decoration (SCD) parameters are compared for their correlations with isolated-chain radii of gyration (Rgs) and upper critical solution temperatures (UCSTs) of polyampholytes modeled by random phase approximation, field-theoretic simulation, and coarse-grained molecular dynamics. SCD is superior to κ in predicting Rg because SCD accounts for effects of contact order, i.e., nonlocality, on dimensions of isolated chains. In contrast, κ and SCD are comparably good, though nonideal, predictors of UCST because frequencies of interchain contacts in the multiple-chain condensed phase are less sensitive to sequence positions than frequencies of intrachain contacts of an isolated chain, as reflected by κ correlating better with condensed-phase interaction energy than SCD.

Review and Comparative Analysis of Methods and Advancements in Predicting Protein Complex Structure.

Protein complexes perform diverse biological functions, and obtaining their three-dimensional structure is critical to understanding and grasping their functions. In many cases, it's not just two proteins interacting to form a dimer; instead, multiple proteins interact to form a multimer. Experimentally resolving protein complex structures can be quite challenging. Recently, there have been efforts and methods that build upon prior predictions of dimer structures to attempt to predict multimer structures. However, in comparison to monomeric protein structure prediction, the accuracy of protein complex structure prediction remains relatively low. This paper provides an overview of recent advancements in efficient computational models for predicting protein complex structures. We introduce protein-protein docking methods in detail and summarize their main ideas, applicable modes, and related information. To enhance prediction accuracy, other critical protein-related information is also integrated, such as predicting interchain residue contact, utilizing experimental data like cryo-EM experiments, and considering protein interactions and non-interactions. In addition, we comprehensively reviewcomputational approaches for end-to-end prediction ofprotein complex structuresbased on artificial intelligence (AI) technologyand describe commonly used datasets and representative evaluation metrics in protein complexes. Finally, we analyze the formidable challenges faced in current protein complex structure prediction tasks, including the structure prediction of heteromeric complex, disordered regions in complex, antibody-antigen complex, and RNA-related complex, as well as the evaluation metrics for complex assessment. We hope that this work will provide comprehensive knowledge of complex structure predictions to contribute to future advanced predictions.

Direct computations of viscoelastic moduli of biomolecular condensates.

In vitro facsimiles of biomolecular condensates are formed by different types of intrinsically disordered proteins including prion-like low complexity domains (PLCDs). PLCD condensates are viscoelastic materials defined by time-dependent, sequence-specific complex shear moduli. Here, we show that viscoelastic moduli can be computed directly using a generalization of the Rouse model and information regarding intra- and inter-chain contacts that is extracted from equilibrium configurations of lattice-based Metropolis Monte Carlo (MMC) simulations. The key ingredient of the generalized Rouse model is the Zimm matrix that we compute from equilibrium MMC simulations. We compute two flavors of Zimm matrices, one referred to as the single-chain model that accounts only for intra-chain contacts, and the other referred to as a collective model, that accounts for inter-chain interactions. The single-chain model systematically overestimates the storage and loss moduli, whereas the collective model reproduces the measured moduli with greater fidelity. However, in the long time, low-frequency domain, a mixture of the two models proves to be most accurate. In line with the theory of Rouse, we find that a continuous distribution of relaxation times exists in condensates. The single crossover frequency between dominantly elastic versus dominantly viscous behaviors is influenced by the totality of the relaxation modes. Hence, our analysis suggests that viscoelastic fluid-like condensates are best described as generalized Maxwell fluids. Finally, we show that the complex shear moduli can be used to solve an inverse problem to obtain distributions of relaxation times that underlie the dynamics within condensates.

Charge-driven complexes formed by conjugated polyelectrolytes as sensing platforms for biogenic amines

We developed sensing platforms to detect and differentiate various types of biogenic amines, cadaverine, putrescine, spermine, histamine, tyramine and tryptamine by monitoring the changes in fluorescence signal. The sensing platforms comprised a carboxylated polyfluorene derivative doped with benzothiadiazole acceptor moieties in a small fraction (20%). When utilized as an individual sensing component, the conjugated polyelectrolyte detects all BAs nonselectively via binding-induced fluorescence quenching. If the conjugated polyelectrolyte forms an intermediary charge-driven complex with an oppositely charged surfactant, a Förster resonance energy transfer (FRET) response is triggered. Biogenic amines are detected through their effects on the FRET signal as enhancement, reduction, or no effect. We formed intermediary charge-driven complexes at nonstoichiometric, cation deficient and excess conditions. Supramolecular interactions governed the co-assembly and disassembly caused by the targeted amines on the conjugated polyelectrolyte-surfactant complex. Therefore, structural and charge-related differences enabled their selective detection in our sensing platforms. Among them, spermine exhibited the highest FRET enhancement effect due to its polycationic aliphatic structure by engaging in multi-point interchain contacts among conjugated polyelectrolyte chains. In contrast, tryptamine showed a unique quenching effect on the FRET signal due to the steric hindrance created by its bulky, heterocyclic structure, leading to the dissociation of the complexes.

Conjugated Polyelectrolyte Thin Films for Pseudocapacitive Applications.

A subclass of organic semiconductors known as conjugated polyelectrolytes (CPEs) is characterized by a conjugated backbone with ionic pendant groups. The water solubility of CPEs typically hinders applications of thin films in aqueous media. Herein, it is reported that films of an anionic CPE, namely CPE-K, drop cast from water produces single-component solid-state pseudocapacitive electrodes that are insoluble in aqueous electrolyte. That X-ray diffraction experiments reveal a more structurally ordered film, relative to the as-obtained powder from chemical synthesis, and dynamic light scattering measurements show an increase in aggregate particle size with increasing [KCl] indicate that CPE-K films are insoluble because of tight interchain contacts and electrostatic screening by the electrolyte. CPE-K film electrodes can maintain 85% of their original capacitance (84 F g-1 ) at 500 A g-1 and exhibit excellent cycling stability, where a capacitance retention of 93% after 100000 cycles at a current density of 35 A g-1 . These findings demonstrate that it is possible to use initially water soluble ionic-organic materials in aqueous electrolytes, by increasing the electrolyte concentration. This strategy can be applied to the application of conjugated polyelectrolytes in batteries, organic electrochemical transistors, and electrochemical sensors, where fast electron and ion transport are required.

Solution NMR assignments and structure for the dimeric kinesin neck domain.

Kinesin is a motor protein, comprised of two heavy and two light chains that transports cargo along the cytoskeletal microtubule filament network. The heavy chain has a neck domain connecting the ATPase motor head responsible for walking along microtubules, with the stalk and subsequent tail domains that bind cargo. The neck domain consists of a coiled coli homodimer with about five heptad repeats, preceded by a linker region that joins to the ATPase head. Here we report 1H, 15N, and 13C NMR assignments and a solution structure for the kinesin neck domain from rat isoform Kif5c. The calculation of the NMR structure of the homodimer was facilitated by unambiguously assigning sidechain NOEs between heptad a and d positions to interchain contacts, since these positions are too far apart to give sidechain contacts in the monomers. The dimeric coiled coil NMR structure is similar to the previously described X-ray structure, whereas the linker region is disordered in solution but contains a short segment with β-strand propensity- the β-linker. Only the coiled coil is protected from solvent exchange, with ∆G values for hydrogen exchange on the order of 4-6kcal/mol. The high stability of the hydrogen-bonded α-helical structure makes it unlikely that unzippering of the coiled coil is involved in kinesin walking. Rather, the linker region serves as a flexible hinge between the kinesin head and neck.

Integrating unsupervised language model with multi-view multiple sequence alignments for high-accuracy inter-chain contact prediction

Accurate identification of inter-chain contacts in the protein complex is critical to determine the corresponding 3D structures and understand the biological functions. We proposed a new deep learning method, ICCPred, to deduce the inter-chain contacts from the amino acid sequences of the protein complex. This pipeline was built on the designed deep residual network architecture, integrating the pre-trained language model with three multiple sequence alignments (MSAs) from different biological views. Experimental results on 709 non-redundant benchmarking protein complexes showed that the proposed ICCPred significantly increased inter-chain contact prediction accuracy compared to the state-of-the-art approaches. Detailed data analyses showed that the significant advantage of ICCPred lies in the utilization of pre-trained transformer language models which can effectively extract the complementary co-evolution diversity from three MSAs. Meanwhile, the designed deep residual network enhances the correlation between the co-evolution diversity and the patterns of inter-chain contacts. These results demonstrated a new avenue for high-accuracy deep-learning inter-chain contact prediction that is applicable to large-scale protein-protein interaction annotations from sequence alone.

Deep transfer learning for inter-chain contact predictions of transmembrane protein complexes

Membrane proteins are encoded by approximately a quarter of human genes. Inter-chain residue-residue contact information is important for structure prediction of membrane protein complexes and valuable for understanding their molecular mechanism. Although many deep learning methods have been proposed to predict the intra-protein contacts or helix-helix interactions in membrane proteins, it is still challenging to accurately predict their inter-chain contacts due to the limited number of transmembrane proteins. Addressing the challenge, here we develop a deep transfer learning method for predicting inter-chain contacts of transmembrane protein complexes, named DeepTMP, by taking advantage of the knowledge pre-trained from a large data set of non-transmembrane proteins. DeepTMP utilizes a geometric triangle-aware module to capture the correct inter-chain interaction from the coevolution information generated by protein language models. DeepTMP is extensively evaluated on a test set of 52 self-associated transmembrane protein complexes, and compared with state-of-the-art methods including DeepHomo2.0, CDPred, GLINTER, DeepHomo, and DNCON2_Inter. It is shown that DeepTMP considerably improves the precision of inter-chain contact prediction and outperforms the existing approaches in both accuracy and robustness.

Optimizing the Martini 3 Force Field Reveals the Effects of the Intricate Balance between Protein-Water Interaction Strength and Salt Concentration on Biomolecular Condensate Formation.

Condensation/dissolution has become a widely acknowledged biological macromolecular assembly phenomenon in subcellular compartmentalization. The MARTINI force field offers a coarse-grained protein model with a resolution that preserves molecular details with an explicit (CG) solvent. Despite its relatively higher resolution, it can still achieve condensate formation in a reasonable computing time with explicit solvent and ionic species. Therefore, it is highly desirable to tune this force field to be able to reproduce the experimentally observed properties of the condensate formation. In this work, we studied the condensate formation of the low-sequence complexity domain of fused in sarcoma protein using a MARTINI 3 force field by systematically modifying (increasing) the protein-water interaction strength and varying the salt concentration. We found that the condensate formation is sensitive both to the protein-water interaction strength and the presence of salt. While the unmodified MARTINI force field yields a complete collapse of proteins into one dense phase (i.e., no dilute phase), we reported a range of modified protein-water interaction strength that is capable of capturing the experimentally found transfer free energy between dense and dilute phases. We also found that the condensates lose their spherical shape upon the addition of salt, especially when the protein-water interactions are weak. Interchain amino acid contact map analysis showed one explanation for this observation: the protein-protein contact fraction reduces as salt is added to systems (when the protein-water interactions are weak), consistent with electrostatic screening effects. This reduction might be responsible for the condensates becoming nonspherical upon the addition of salt by reducing the need for minimizing the interfacial area. However, as the protein-water interactions become stronger to the extent that makes the transfer free energy agree well with experimentally observed transfer free energy, we found an increase in the protein-protein contact fraction upon the addition of salt, consistent with the salting-out effects. Therefore, we concluded that there is an intricate balance between screening effects and salting-out effects upon the addition of salt and this balance is highly sensitive to the strength of protein-water interactions.

Selective laser sintering of functionalized carbon nanotubes and inorganic fullerene-like tungsten disulfide reinforced polyamide 12 nanocomposites with excellent fire safety and mechanical properties

More effective and safe fire prevention solutions for polymer derivatives are necessary to address rising environmental and health concerns. In this study, the novel and simple chemically modified multi-walled carbon nanotubes (H-CNT) and PEGylated inorganic fullerene tungsten sulphide (P-IF-WS2) reinforced PA12 nanocomposites are synergistically prepared via selective laser sintering (SLS). The PA12/H-CNT/IF-WS2 nanocomposites show a wider sintering window than blank PA12. The formed hydrogen bond network of inter-chain contacts and compatibility in PA12/H-CNT/P-IF-WS2 matrix is proved. Furthermore, a dense carbon layer is developed by further dehydrating the oxygen-rich functional groups on the surface of H-CNT and P-IF-WS2 during carbonization at high temperatures. The PA12/H-CNT/P-IF-WS2 nanocomposites fabricated by incorporating hydrogen and carbon bonds in the PA12-based nanocomposites demonstrate good fire safety, thermal, and mechanical properties. The significant reduction in total heat release (20.14%), peak heat release (38.9%), and total smoke emission (22.6%) showed the improved fire safety of PA12. The H-CNT and P-IF-WS2 nanofillers also enhanced the mechanical (tensile and dynamic mechanical) capabilities. This technique of introducing nano-additives to SLS samples changes offers a practical, long-lasting, and effective way to improve the flame retardancy of laser-sintered polymer nanocomposites.

Extension and long-range contacts drive chromatin clutches and aggregates.

Chromatin organization plays a vital role in gene expression and DNA replication. Whether the local arrangement in vivo is a 10-nm array or 30-nm fiber remains controversial. Moreover, it remains unclear whether the structure and dynamics are closer to a liquid or solid on a larger scale. Multiple factors, such as DNA linker lengths, chromatin remodelers, histone modifications, and ionic environments, can affect the structure of chromatin. Using a near-atomistic coarse-grained model without additional assumptions, we studied the force-extension behavior of uniform canonical chromatin with 12 nucleosomes and 20-bp linkers. We found that forces below 3 pN result in a spring-like linear extension as nucleosomes undergo shear motions. However, forces beyond 3.5 pN break chromatin into trimer and tetramer clutches, thus enabling a dramatic non-uniform extension. Meanwhile, inter-chain contacts and interdigitation can facilitate chromatin unfolding. Under a crowding environment, chromatin unfolds and exposes more nucleosomal surfaces, allowing favorable inter-chain inter-nucleosomal interactions, such as stackings. These results reconcile observed ordered fibers in vitro and disordered arrays in vivo. The high nucleosome concentration and the tension provided by molecular motors can cause disordered chromatin arrays to prevail in vivo. These factors may be relevant to the irregularity of heterochromatin and mitotic chromosomes. Furthermore, the strong inter-chain contacts can lead to the formation of condensates. By applying such models under different nucleosome and salt concentrations, we will provide more insights into whether chromatin favors liquid- or solid-like states. We can also test whether long-range contacts can drive percolation transitions. Our simulations provide a detailed microscopic description of genome organization and characterize chromatin's local and bulk properties.

Unraveling the phase separation of EWS low complexity domain using molecular dynamics simulations.

EWS, an RNA-binding protein, is responsible for gene expression, cell signaling, RNA processing, and transport. It is a member of the FET (FUS, EWS, TAF15) family, whose derailed phase separation is known to cause many neurodegenerative diseases. Through achromosomal translocation, the low complexity domain (LCD) of EWS may be fused to the DNA binding domain of the transcription factor FLI1. The resultant EWS-FLI1 fusion is a potent oncogenic protein responsible for the development and growth of Ewing sarcoma and aggressive bone cancer, often affecting adolescents and young adults. Though studies have shown that the LCD is capable of liquid-liquid phase separation (LLPS), mechanistic details are still lacking. In this study, using molecular dynamics simulations, we aim to understand at atomic resolution the mechanism underpinning phase separation of the EWS-LCD. We have performed single-chain all-atom simulations and multi-chain coarse-grain simulations to investigate the intra and interchain contacts responsible for LLPS of EWS-LCD. Our study reveals that Tyrosine and Glutamine residues form contacts with the greatest frequency in both single and multi-chain interactions. We observed that residues near the C-terminus are more likely to form contacts than N-terminal residues. Our results are consistent with experimental results derived from biophysical investigations. This study is a step towards understanding the self-association of the EWS-LCD at the atomic level, and it will provide insight into the specific interactions that drive EWS phase separation.

Prediction of inter-chain distance maps of protein complexes with 2D attention-based deep neural networks

Residue-residue distance information is useful for predicting tertiary structures of protein monomers or quaternary structures of protein complexes. Many deep learning methods have been developed to predict intra-chain residue-residue distances of monomers accurately, but few methods can accurately predict inter-chain residue-residue distances of complexes. We develop a deep learning method CDPred (i.e., Complex Distance Prediction) based on the 2D attention-powered residual network to address the gap. Tested on two homodimer datasets, CDPred achieves the precision of 60.94% and 42.93% for top L/5 inter-chain contact predictions (L: length of the monomer in homodimer), respectively, substantially higher than DeepHomo’s 37.40% and 23.08% and GLINTER’s 48.09% and 36.74%. Tested on the two heterodimer datasets, the top Ls/5 inter-chain contact prediction precision (Ls: length of the shorter monomer in heterodimer) of CDPred is 47.59% and 22.87% respectively, surpassing GLINTER’s 23.24% and 13.49%. Moreover, the prediction of CDPred is complementary with that of AlphaFold2-multimer.

Identifying the Template for Oligomer to Fibril Conversion for Amyloid-β (1-42) Oligomers using Hamiltonian Replica Exchange Molecular Dynamics.

The toxicity of amyloid-β (Aβ) oligomers has been known to be higher compared to mature fibrils. Yet the presence of plaques in Alzheimer's disease patients indicates the significance of oligomer to fibril conversion for Aβ aggregates. In this study, we investigate Aβ13-42 oligomers having two to five peptide chains using extensive all-atom molecular dynamics simulations to identify the on- or off-pathway intermediates in fibril formation pathway. Hamiltonian replica exchange method through solute tempering (REST2) has been employed to explore the different structures attained by these aggregates. Using intra-chain and inter-chain contacts as reaction coordinates, we obtain the free energy surface for the Aβ13-42 oligomers. Consequently, their stable conformations and structural features have been identified. The found conformations belonging to most probable structures possess both parallel and anti-parallel β-sheets, characteristic of on- and off-pathway intermediates, respectively. Further, we have measured the tendency to form fibril like interactions among the β-sheet forming residues. Our analysis finds that residues 30-36 possess higher tendency to form fibril like contacts among all the residues. While we find stronger interaction among residues 30-36, these amino acids are also found to be more shielded from water compared to others. With previous experimental studies finding these residues to be more crucial for the stability of Aβ42 oligomers, we propose that interactions within this patch could trigger seed formation that leads to conversion of on-pathway oligomers into disease relevant fibrils.

A deep dilated convolutional residual network for predicting interchain contacts of protein homodimers.

MotivationDeep learning has revolutionized protein tertiary structure prediction recently. The cutting-edge deep learning methods such as AlphaFold can predict high-accuracy tertiary structures for most individual protein chains. However, the accuracy of predicting quaternary structures of protein complexes consisting of multiple chains is still relatively low due to lack of advanced deep learning methods in the field. Because interchain residue–residue contacts can be used as distance restraints to guide quaternary structure modeling, here we develop a deep dilated convolutional residual network method (DRCon) to predict interchain residue–residue contacts in homodimers from residue–residue co-evolutionary signals derived from multiple sequence alignments of monomers, intrachain residue–residue contacts of monomers extracted from true/predicted tertiary structures or predicted by deep learning, and other sequence and structural features.ResultsTested on three homodimer test datasets (Homo_std dataset, DeepHomo dataset and CASP-CAPRI dataset), the precision of DRCon for top L/5 interchain contact predictions (L: length of monomer in a homodimer) is 43.46%, 47.10% and 33.50% respectively at 6 Å contact threshold, which is substantially better than DeepHomo and DNCON2_inter and similar to Glinter. Moreover, our experiments demonstrate that using predicted tertiary structure or intrachain contacts of monomers in the unbound state as input, DRCon still performs well, even though its accuracy is lower than using true tertiary structures in the bound state are used as input. Finally, our case study shows that good interchain contact predictions can be used to build high-accuracy quaternary structure models of homodimers.Availability and implementationThe source code of DRCon is available at https://github.com/jianlin-cheng/DRCon. The datasets are available at https://zenodo.org/record/5998532#.YgF70vXMKsB.Supplementary information Supplementary data are available at Bioinformatics online.

Emerging mutation patterns in SARS-CoV-2 variants

There is an urgent need to understand the functional effects of mutations in emerging variants of SARS-CoV-2. Variants of concern (alpha, beta, gamma and delta) acquired four patterns of spike glycoprotein mutations that enhance transmissibility and immune evasion: 1) mutations in the N-terminal domain (NTD), 2) mutations in the Receptor Binding Domain (RBD), 3) mutations at interchain contacts of the spike trimer, and 4) furin cleavage site mutations. Most distinguishing mutations among variants of concern are exhibited in the NTD, localized to sites of high structural flexibility. Emerging variants of interest such as mu, lambda and C.1.2 exhibit the same patterns of mutations as variants of concern. There is a strong likelihood that SARS-CoV-2 variants will continue to emerge with mutations in these defined patterns, thus providing a basis for the development of next line antiviral drugs and vaccine candidates.