ABSTRACT
Molecular dynamics simulation is a crucial research domain within the life sciences, focusing on comprehending the mechanisms of biomolecular interactions at atomic scales. Protein simulation, as a critical subfield, often utilizes MD for implementation, with trajectory data play a pivotal role in drug discovery. The advancement of high-performance computing and deep learning technology becomes popular and critical to predict protein properties from vast trajectory data, posing challenges regarding data features extraction from the complicated simulation data and dimensionality reduction. Simultaneously, it is essential to provide a meaningful explanation of the biological mechanism behind dimensionality. To tackle this challenge, we propose a new unsupervised model named RevGraphVAMP to intelligently analyze the simulation trajectory. This model is based on the variational approach for Markov processes (VAMP) and integrates graph convolutional neural networks and physical constraint optimization to enhance the learning performance. Additionally, we introduce attention mechanism to assess the importance of key interaction region, facilitating the interpretation of molecular mechanism. In comparison to other VAMPNets models, our model showcases competitive performance, improved accuracy in state transition prediction, as demonstrated through its application to two public datasets and the Shank3-Rap1 complex, which is associated with autism spectrum disorder. Moreover, it enhanced dimensionality reduction discrimination across different substates and provides interpretable results for protein structural characterization.
Subject(s)
Markov Chains , Molecular Dynamics Simulation , Neural Networks, Computer , Proteins , Proteins/chemistry , Humans , Deep LearningABSTRACT
The automatic detection of cells in microscopy image sequences is a significant task in biomedical research. However, routine microscopy images with cells, which are taken during the process whereby constant division and differentiation occur, are notoriously difficult to detect due to changes in their appearance and number. Recently, convolutional neural network (CNN)-based methods have made significant progress in cell detection and tracking. However, these approaches require many manually annotated data for fully supervised training, which is time-consuming and often requires professional researchers. To alleviate such tiresome and labor-intensive costs, we propose a novel weakly supervised learning cell detection and tracking framework that trains the deep neural network using incomplete initial labels. Our approach uses incomplete cell markers obtained from fluorescent images for initial training on the Induced Pluripotent Stem (iPS) cell dataset, which is rarely studied for cell detection and tracking. During training, the incomplete initial labels were updated iteratively by combining detection and tracking results to obtain a model with better robustness. Our method was evaluated using two fields of the iPS cell dataset, along with the cell detection accuracy (DET) evaluation metric from the Cell Tracking Challenge (CTC) initiative, and it achieved 0.862 and 0.924 DET, respectively. The transferability of the developed model was tested using the public dataset FluoN2DH-GOWT1, which was taken from CTC; this contains two datasets with reference annotations. We randomly removed parts of the annotations in each labeled data to simulate the initial annotations on the public dataset. After training the model on the two datasets, with labels that comprise 10% cell markers, the DET improved from 0.130 to 0.903 and 0.116 to 0.877. When trained with labels that comprise 60% cell markers, the performance was better than the model trained using the supervised learning method. This outcome indicates that the model's performance improved as the quality of the labels used for training increased.
Subject(s)
Neural Networks, Computer , Supervised Machine Learning , Image Processing, Computer-Assisted/methodsABSTRACT
Recent studies have shed light on the potential of circular RNA (circRNA) as a biomarker for disease diagnosis and as a nucleic acid vaccine. The exploration of these functionalities requires correct circRNA full-length sequences; however, existing assembly tools can only correctly assemble some circRNAs, and their performance can be further improved. Here, we introduce a novel feature known as the junction contig (JC), which is an extension of the back-splice junction (BSJ). Leveraging the strengths of both BSJ and JC, we present a novel method called JCcirc (https://github.com/cbbzhang/JCcirc). It enables efficient reconstruction of all types of circRNA full-length sequences and their alternative isoforms using splice graphs and fragment coverage. Our findings demonstrate the superiority of JCcirc over existing methods on human simulation datasets, and its average F1 score surpasses CircAST by 0.40 and outperforms both CIRI-full and circRNAfull by 0.13. For circRNAs below 400 bp, 400-800 bp, 800 bp-1200 bp and above 1200 bp, the correct assembly rates are 0.13, 0.09, 0.04 and 0.03 higher, respectively, than those achieved by existing methods. Moreover, JCcirc also outperforms existing assembly tools on other five model species datasets and real sequencing datasets. These results show that JCcirc is a robust tool for accurately assembling circRNA full-length sequences, laying the foundation for the functional analysis of circRNAs.
Subject(s)
RNA, Circular , RNA , Humans , RNA, Circular/genetics , Sequence Analysis, RNA/methods , Protein Isoforms/genetics , RNA/geneticsABSTRACT
Darobactin is a heptapeptide antibiotic featuring an ether cross-link and a C-C cross-link, and both cross-links are installed by a radical S-adenosylmethionine (rSAM) enzyme DarE. How a single DarE enzyme affords the two chemically distinct cross-links remains largely obscure. Herein, by mapping the biosynthetic landscape for darobactin-like RiPP (daropeptide), we identified and characterized two novel daropeptides that lack the C-C cross-link present in darobactin and instead are solely composed of ether cross-links. Phylogenetic and mutagenesis analyses reveal that the daropeptide maturases possess intrinsic multifunctionality, catalyzing not only the formation of ether cross-link but also C-C cross-linking and Ser oxidation. Intriguingly, the different chemical outcomes are controlled by the exact substrate motifs. Our work not only provides a roadmap for the discovery of new daropeptide natural products but also offers insights into the regulatory mechanisms that govern these remarkably versatile ether cross-link-forming rSAM enzymes.
Subject(s)
Ether , S-Adenosylmethionine , S-Adenosylmethionine/chemistry , Phylogeny , Ethers , Ethyl Ethers , CatalysisABSTRACT
Drivers who exhibit dangerous driving behaviours, such as aggressive, risky, and negative emotion cognition driving, are more likely to be involved in road crashes. A key motivator behind unsafe driving behaviours is driving anger. However, it is unclear whether lifestyle, driving anger, and dangerous driving behaviours are related. A total of 344 Chinese drivers with a formal driving license were asked to complete the socio-demographic information, the Chinese lifestyle questionnaire (Self-designed), the 14 items Driving Anger Scale (DAS), and the Dula Dangerous Driving Index (DDDI). The Chinese driver's lifestyles were analysed using Exploratory Factor Analysis (EFA), revealing a four-factor structure ("Culture", "Workaholism", "Sports" and "Amusement"). The 14 items DAS factor structure was determined using a Confirmatory Factor Analysis (CFA), yielding a two-factor structure ("Safety Concern anger" and "Arrival Concern anger"). Based on Hierarchical Multiple Regression (HMR), only "Workaholism" was associated with aggressive, risky, and negative emotion cognition driving. The trait driving anger was examined as a mediator between the "Workaholism" and dangerous driving (aggressive, risky, and negative emotion cognition driving) through a Structural Equation Modelling (SEM) approach. "Workaholism" was shown to influence these dangerous driving behaviours through trait driving anger. Lastly, this article discussed the theoretical and practical implications and research limitations.
ABSTRACT
The development of noninvasive brain imaging such as resting-state functional magnetic resonance imaging (rs-fMRI) and its combination with AI algorithm provides a promising solution for the early diagnosis of Autism spectrum disorder (ASD). However, the performance of the current ASD classification based on rs-fMRI still needs to be improved. This paper introduces a classification framework to aid ASD diagnosis based on rs-fMRI. In the framework, we proposed a novel filter feature selection method based on the difference between step distribution curves (DSDC) to select remarkable functional connectivities (FCs) and utilized a multilayer perceptron (MLP) which was pretrained by a simplified Variational Autoencoder (VAE) for classification. We also designed a pipeline consisting of a normalization procedure and a modified hyperbolic tangent (tanh) activation function to replace the classical tanh function, further improving the model accuracy. Our model was evaluated by 10 times 10-fold cross-validation and achieved an average accuracy of 78.12%, outperforming the state-of-the-art methods reported on the same dataset. Given the importance of sensitivity and specificity in disease diagnosis, two constraints were designed in our model which can improve the model's sensitivity and specificity by up to 9.32% and 10.21%, respectively. The added constraints allow our model to handle different application scenarios and can be used broadly.
Subject(s)
Autism Spectrum Disorder , Brain , Brain Mapping , Humans , Magnetic Resonance Imaging , Neural Networks, ComputerABSTRACT
Residue distance prediction from the sequence is critical for many biological applications such as protein structure reconstruction, protein-protein interaction prediction, and protein design. However, prediction of fine-grained distances between residues with long sequence separations still remains challenging. In this study, we propose DuetDis, a method based on duet feature sets and deep residual network with squeeze-and-excitation (SE), for protein inter-residue distance prediction. DuetDis embraces the ability to learn and fuse features directly or indirectly extracted from the whole-genome/metagenomic databases and, therefore, minimize the information loss through ensembling models trained on different feature sets. We evaluate DuetDis and 11 widely used peer methods on a large-scale test set (610 proteins chains). The experimental results suggest that 1) prediction results from different feature sets show obvious differences; 2) ensembling different feature sets can improve the prediction performance; 3) high-quality multiple sequence alignment (MSA) used for both training and testing can greatly improve the prediction performance; and 4) DuetDis is more accurate than peer methods for the overall prediction, more reliable in terms of model prediction score, and more robust against shallow multiple sequence alignment (MSA).
ABSTRACT
Bioinformatics analysis has been playing a vital role in identifying potential genomic biomarkers more accurately from an enormous number of candidates by reducing time and cost compared to the wet-lab-based experimental procedures for disease diagnosis, prognosis, and therapies. Cervical cancer (CC) is one of the most malignant diseases seen in women worldwide. This study aimed at identifying potential key genes (KGs), highlighting their functions, signaling pathways, and candidate drugs for CC diagnosis and targeting therapies. Four publicly available microarray datasets of CC were analyzed for identifying differentially expressed genes (DEGs) by the LIMMA approach through GEO2R online tool. We identified 116 common DEGs (cDEGs) that were utilized to identify seven KGs (AURKA, BRCA1, CCNB1, CDK1, MCM2, NCAPG2, and TOP2A) by the protein-protein interaction (PPI) network analysis. The GO functional and KEGG pathway enrichment analyses of KGs revealed some important functions and signaling pathways that were significantly associated with CC infections. The interaction network analysis identified four TFs proteins and two miRNAs as the key transcriptional and post-transcriptional regulators of KGs. Considering seven KGs-based proteins, four key TFs proteins, and already published top-ranked seven KGs-based proteins (where five KGs were common with our proposed seven KGs) as drug target receptors, we performed their docking analysis with the 80 meta-drug agents that were already published by different reputed journals as CC drugs. We found Paclitaxel, Vinorelbine, Vincristine, Docetaxel, Everolimus, Temsirolimus, and Cabazitaxel as the top-ranked seven candidate drugs. Finally, we investigated the binding stability of the top-ranked three drugs (Paclitaxel, Vincristine, Vinorelbine) by using 100 ns MD-based MM-PBSA simulations with the three top-ranked proposed receptors (AURKA, CDK1, TOP2A) and observed their stable performance. Therefore, the proposed drugs might play a vital role in the treatment against CC.
Subject(s)
Computational Biology , Uterine Cervical Neoplasms , Aurora Kinase A/genetics , Biomarkers, Tumor/genetics , Chromosomal Proteins, Non-Histone/genetics , Computational Biology/methods , Databases, Genetic , Early Detection of Cancer/methods , Female , Gene Expression Profiling/methods , Gene Expression Regulation, Neoplastic , Gene Regulatory Networks , Humans , Paclitaxel , RNA, Messenger , Uterine Cervical Neoplasms/drug therapy , Uterine Cervical Neoplasms/genetics , Vincristine , VinorelbineABSTRACT
Sequence-based residue contact prediction plays a crucial role in protein structure reconstruction. In recent years, the combination of evolutionary coupling analysis (ECA) and deep learning (DL) techniques has made tremendous progress for residue contact prediction, thus a comprehensive assessment of current methods based on a large-scale benchmark data set is very needed. In this study, we evaluate 18 contact predictors on 610 non-redundant proteins and 32 CASP13 targets according to a wide range of perspectives. The results show that different methods have different application scenarios: (1) DL methods based on multi-categories of inputs and large training sets are the best choices for low-contact-density proteins such as the intrinsically disordered ones and proteins with shallow multi-sequence alignments (MSAs). (2) With at least 5L (L is sequence length) effective sequences in the MSA, all the methods show the best performance, and methods that rely only on MSA as input can reach comparable achievements as methods that adopt multi-source inputs. (3) For top L/5 and L/2 predictions, DL methods can predict more hydrophobic interactions while ECA methods predict more salt bridges and disulfide bonds. (4) ECA methods can detect more secondary structure interactions, while DL methods can accurately excavate more contact patterns and prune isolated false positives. In general, multi-input DL methods with large training sets dominate current approaches with the best overall performance. Despite the great success of current DL methods must be stated the fact that there is still much room left for further improvement: (1) With shallow MSAs, the performance will be greatly affected. (2) Current methods show lower precisions for inter-domain compared with intra-domain contact predictions, as well as very high imbalances in precisions between intra-domains. (3) Strong prediction similarities between DL methods indicating more feature types and diversified models need to be developed. (4) The runtime of most methods can be further optimized.
Subject(s)
Computational Biology/methods , Proteins/chemistry , Amino Acid Sequence , Datasets as Topic , Deep Learning , Prospective Studies , Retrospective StudiesABSTRACT
Understanding the conformational dynamics of proteins and peptides involved in important functions is still a difficult task in computational structural biology. Because such conformational transitions in ß-amyloid (Aß) forming peptides play a crucial role in many neurological disorders, researchers from different scientific fields have been trying to address issues related to the folding of Aß forming peptides together. Many theoretical models have been proposed in the recent years for studying Aß peptides using mathematical, physicochemical, and molecular dynamics simulation, and machine learning approaches. In this article, we have comprehensively reviewed the developmental advances in the theoretical models for Aß peptide folding and interactions, particularly in the context of neurological disorders. Furthermore, we have extensively reviewed the advances in molecular dynamics simulation as a tool used for studying the conversions between polymorphic amyloid forms and applications of using machine learning approaches in predicting Aß peptides and aggregation-prone regions in proteins. We have also provided details on the theoretical advances in the study of Aß peptides, which would enhance our understanding of these peptides at the molecular level and eventually lead to the development of targeted therapies for certain acute neurological disorders such as Alzheimer's disease in the future.
ABSTRACT
We consider the effect of lauric acid on the stability of various fibril-like assemblies of Aß peptides. For this purpose, we have performed molecular dynamics simulations of these assemblies either in complex with lauric acid or without presence of the ligand. While we do not observe a stabilizing effect on Aß40 -fibrils, we find that addition of lauric acid strengthens the stability of fibrils built from the triple-stranded S-shaped Aß42 -peptides considered to be more toxic. Or results may help to understand how the specifics of the brain-environment modulate amyloid formation and propagation.
Subject(s)
Amyloid beta-Peptides/chemistry , Fatty Acids/chemistry , Peptide Fragments/chemistry , Humans , Models, Molecular , Protein StabilityABSTRACT
Peptides build from D-amino acids resist enzymatic degradation. The resulting extended time of biological activity makes them prime candidates for the development of pharmaceuticals. Of special interest are D-retro-inverso (DRI) peptides where a reversed sequence of D-amino acids leads to molecules with almost the same structure, stability, and bioactivity as the parent L-peptides but increased resistance to proteolytic degradation. Here, we study the effect of DRI-Aß40 and DRI-Aß42 peptides on fibril formation. Using molecular dynamics simulations, we compare the stability of typical amyloid fibril models with such where the L-peptides are replaced by DRI-Aß40 and DRI-Aß42 peptides. We then explore the likelihood for cross fibrilization of Aß L- and DRI-peptides by investigating how the presence of DRI peptides alters the elongation and stability of L-Aß-fibrils. Our data suggest that full-length DRI-peptides may enhance the fibril formation and decrease the ratio of soluble toxic Aß oligomers, pointing out potential for D-amino-acid-based drug design targeting Alzheimer's disease.
Subject(s)
Amino Acids/chemistry , Amyloid/chemical synthesis , Molecular Dynamics Simulation , Alzheimer Disease/drug therapy , Alzheimer Disease/metabolism , Amino Acids/metabolism , Amyloid/antagonists & inhibitors , Amyloid/chemistry , Drug Design , Humans , Solubility , StereoisomerismABSTRACT
As the primary toxic species in the etiology of Alzheimer disease (AD) are low molecular weight oligomers of Aß, it is crucial to understand the structure of Aß oligomers for gaining molecular insights into AD pathology. We have earlier demonstrated that in the presence of fatty acids, Aß42 peptides assemble as 12-24mer oligomers. These Large Fatty Acid-derived Oligomers (LFAOs) exist predominantly as 12mers at low and as 24mers at high concentrations. The 12mers are more neurotoxic than the 24mers and undergo self-replication, while the latter propagate to morphologically distinct fibrils with succinct pathological consequences. In order to glean into their functional differences and similarities, we have determined their structures in greater detail by combining molecular dynamic simulations with biophysical measurements. We conjecture that the LFAO are made of Aß units in an S-shaped conformation, with the 12mers forming a double-layered hexamer ring (6 × 2) while the structure of 24mers is a double-layered dodecamer ring (12 × 2). A closer inspection of the (6 × 2) and (12 × 2) structures reveals a concentration and pH dependent molecular reorganization in the assembly of 12 to 24mers, which seems to be the underlying mechanism for the observed biophysical and cellular properties of LFAOs.
Subject(s)
Amyloid beta-Peptides/chemistry , Fatty Acids/chemistry , Peptide Fragments/chemistry , Protein Multimerization , Molecular Dynamics Simulation , Protein Structure, QuaternaryABSTRACT
Colonic amyloidosis is the result of overexpression of the serum amyloid A (SAA) protein in inflammatory bowel disease or colon cancer. Crucial for amyloid formation are the first ten N-terminal residues, which in the crystal structure are a part of a 27-residue long helix. Here, we study this 27-residue N-terminal region of SAA by a multiexchange variant of replica exchange molecular dynamics. An ensemble of configurations is observed, dominated by three motifs: the single helix of the crystal structure, a helix-turn-helix configurations, and such where the residues 14-27 are the part of a helix but the first 13 residues form an extended and disordered segment that is prone to aggregation. The single point mutation E9A shifts the equilibrium to the latter motif, indicating the importance of interactions involving this residue for the stability of the SAA protein.
ABSTRACT
Using a variant of Hamilton-replica-exchange, we study for wild type and Iowa mutant Aß40 the conversion between fibrils with antiparallel ß-sheets and such with parallel ß-sheets. We show that wild type and mutant form distinct salt bridges that in turn stabilize different fibril organizations. The conversion between the two fibril forms leads to the release of small aggregates that in the Iowa mutant may shift the equilibrium from fibrils to more toxic oligomers.
Subject(s)
Amyloid beta-Peptides/chemistry , Amyloid beta-Peptides/genetics , Humans , Models, Molecular , Mutation , Protein Structure, SecondaryABSTRACT
We propose a variant of the recently found S-shaped Aß1-42-motif that is characterized by out-of-register C-terminal ß-strands. We show that chains with this structure can form not only fibrils that are compatible with the NMR signals but also barrel-shaped oligomers that resemble the ones formed by the much smaller cylindrin peptides. By running long all-atom molecular dynamics simulations at physiological temperatures with an explicit solvent, we study the stability of these constructs and show that they are plausible models for neurotoxic oligomers. After analyzing the transitions between different assemblies, we suggest a mechanism for amyloid formation in Alzheimer's disease.
Subject(s)
Amyloid beta-Peptides/chemistry , Molecular Dynamics Simulation , Neurotoxins/chemistry , Peptide Fragments/chemistry , Nuclear Magnetic Resonance, BiomolecularABSTRACT
When assembling as fibrils Aß40 peptides can only assume U-shaped conformations while Aß42 can also arrange as S-shaped three-stranded chains. We show that this allows Aß42 peptides to assemble pore-like structures that may explain their higher toxicity. For this purpose, we develop a scalable model of ring-like assemblies of S-shaped Aß1-42 chains and study the stability and structural properties of these assemblies through atomistic molecular dynamics simulations. We find that the proposed arrangements are in size and symmetry compatible with experimentally observed Aß assemblies. We further show that the interior pore in our models allows for water leakage as a possible mechanism of cell toxicity of Aß42 amyloids.
ABSTRACT
We introduce Replica-Exchange-with-Tunneling (RET) simulations as a tool for studies of the conversion between polymorphic amyloids. For the 11-residue amyloid-forming cylindrin peptide we show that this technique allows for a more efficient sampling of the formation and interconversion between fibril-like and barrel-like assemblies. We describe a protocol for optimized analysis of RET simulations that allows us to propose a mechanism for formation and interconversion between various cylindrin assemblies. Especially, we show that an interchain salt bridge between residues K3 and D7 is crucial for formation of the barrel structure.
Subject(s)
Amyloid/chemistry , alpha-Crystallin B Chain/chemistry , Amino Acid Sequence , Humans , Molecular Dynamics Simulation , Protein Conformation , Protein Folding , Protein Structure, SecondaryABSTRACT
Recent experiments suggest that an amino acid sequence encodes not only the native fold of a protein but also other forms that are essential for its function or are important during folding or association. These various forms populate a multifunnel folding and association landscape where mutations, changes in environment, or interaction with other molecules switch between the encoded folds. We introduce replica exchange with tunneling as a way to efficiently simulate switching between distinct folds of proteins and protein aggregates. The correctness and efficiency of our approach are demonstrated in a series of simulations covering a wide range of proteins, from a small 11-residue large designed peptide to two 56-residue large mutants of the A and B domains of protein G.