In Silico Prediction of New Inhibitors for Kirsten Rat Sarcoma G12D Cancer Drug Target Using Machine Learning-Based Virtual Screening, Molecular Docking, and Molecular Dynamic Simulation Approaches.

Single-point mutations in the Kirsten rat sarcoma (KRAS) viral proto-oncogene are the most common cause of human cancer. In humans, oncogenic KRAS mutations are responsible for about 30% of lung, pancreatic, and colon cancers. One of the predominant mutant KRAS G12D variants is responsible for pancreatic cancer and is an attractive drug target. At the time of writing, no Food and Drug Administration (FDA) approved drugs are available for the KRAS G12D mutant. So, there is a need to develop an effective drug for KRAS G12D. The process of finding new drugs is expensive and time-consuming. On the other hand, in silico drug designing methodologies are cost-effective and less time-consuming. Herein, we employed machine learning algorithms such as K-nearest neighbor (KNN), support vector machine (SVM), and random forest (RF) for the identification of new inhibitors against the KRAS G12D mutant. A total of 82 hits were predicted as active against the KRAS G12D mutant. The active hits were docked into the active site of the KRAS G12D mutant. Furthermore, to evaluate the stability of the compounds with a good docking score, the top two complexes and the standard complex (MRTX-1133) were subjected to 200 ns MD simulation. The top two hits revealed high stability as compared to the standard compound. The binding energy of the top two hits was good as compared to the standard compound. Our identified hits have the potential to inhibit the KRAS G12D mutation and can help combat cancer. To the best of our knowledge, this is the first study in which machine-learning-based virtual screening, molecular docking, and molecular dynamics simulation were carried out for the identification of new promising inhibitors for the KRAS G12D mutant.

Abrogation of ORF8-IRF3 binding interface with Carbon nanotube derivatives to rescue the host immune system against SARS-CoV-2 by using molecular screening and simulation approaches.

The COVID-19 pandemic, caused by the SARS-CoV-2 virus, has led to over six million deaths worldwide. In human immune system, the type 1 interferon (IFN) pathway plays a crucial role in fighting viral infections. However, the ORF8 protein of the virus evade the immune system by interacting with IRF3, hindering its nuclear translocation and consequently downregulate the type I IFN signaling pathway. To block the binding of ORF8-IRF3 and inhibit viral pathogenesis a quick discovery of an inhibitor molecule is needed. Therefore, in the present study, the interface between the ORF8 and IRF3 was targeted on a high-affinity carbon nanotube by using computational tools. After analysis of 62 carbon nanotubes by multiple docking with the induced fit model, the top five compounds with high docking scores of - 7.94 kcal/mol, - 7.92 kcal/mol, - 7.28 kcal/mol, - 7.19 kcal/mol and - 7.09 kcal/mol (top hit1-5) were found to have inhibitory activity against the ORF8-IRF3 complex. Molecular dynamics analysis of the complexes revealed the high compactness of residues, stable binding, and strong hydrogen binding network among the ORF8-nanotubes complexes. Moreover, the total binding free energy for top hit1-5 was calculated to be - 43.21 ± 0.90 kcal/mol, - 41.17 ± 0.99 kcal/mol, - 48.85 ± 0.62 kcal/mol, - 43.49 ± 0.77 kcal/mol, and - 31.18 ± 0.78 kcal/mol respectively. These results strongly suggest that the identified top five nanotubes (hit1-5) possess significant potential for advancing and exploring innovative drug therapies. This underscores their suitability for subsequent in vivo and in vitro experiments, marking them as promising candidates worthy of further investigation.

MF-MNER: Multi-models Fusion for MNER in Chinese Clinical Electronic Medical Records.

To address the problem of poor entity recognition performance caused by the lack of Chinese annotation in clinical electronic medical records, this paper proposes a multi-medical entity recognition method F-MNER using a fusion technique combining BART, Bi-LSTM, and CRF. First, after cleaning, encoding, and segmenting the electronic medical records, the obtained semantic representations are dynamically fused using a bidirectional autoregressive transformer (BART) model. Then, sequential information is captured using a bidirectional long short-term memory (Bi-LSTM) network. Finally, the conditional random field (CRF) is used to decode and output multi-task entity recognition. Experiments are performed on the CCKS2019 dataset, with micro avg Precision, macro avg Recall, weighted avg Precision reaching 0.880, 0.887, and 0.883, and micro avg F1-score, macro avg F1-score, weighted avg F1-score reaching 0.875, 0.876, and 0.876 respectively. Compared with existing models, our method outperforms the existing literature in three evaluation metrics (micro average, macro average, weighted average) under the same dataset conditions. In the case of weighted average, the Precision, Recall, and F1-score are 19.64%, 15.67%, and 17.58% higher than the existing BERT-BiLSTM-CRF model respectively. Experiments are performed on the actual clinical dataset with our MF-MNER, the Precision, Recall, and F1-score are 0.638, 0.825, and 0.719 under the micro-avg evaluation mechanism. The Precision, Recall, and F1-score are 0.685, 0.800, and 0.733 under the macro-avg evaluation mechanism. The Precision, Recall, and F1-score are 0.647, 0.825, and 0.722 under the weighted avg evaluation mechanism. The above results show that our method MF-MNER can integrate the advantages of BART, Bi-LSTM, and CRF layers, significantly improving the performance of downstream named entity recognition tasks with a small amount of annotation, and achieving excellent performance in terms of recall score, which has certain practical significance. Source code and datasets to reproduce the results in this paper are available at https://github.com/xfwang1969/MF-MNER .

Identification of an immune-related gene signature for predicting prognosis and immunotherapy efficacy in liver cancer via cell-cell communication.

BACKGROUND: Liver cancer is one of the deadliest malignant tumors worldwide. Immunotherapy has provided hope to patients with advanced liver cancer, but only a small fraction of patients benefit from this treatment due to individual differences. Identifying immune-related gene signatures in liver cancer patients not only aids physicians in cancer diagnosis but also offers personalized treatment strategies, thereby improving patient survival rates. Although several methods have been developed to predict the prognosis and immunotherapeutic efficacy in patients with liver cancer, the impact of cell-cell interactions in the tumor microenvironment has not been adequately considered. AIM: To identify immune-related gene signals for predicting liver cancer prognosis and immunotherapy efficacy. METHODS: Cell grouping and cell-cell communication analysis were performed on single-cell RNA-sequencing data to identify highly active cell groups in immune-related pathways. Highly active immune cells were identified by intersecting the highly active cell groups with B cells and T cells. The significantly differentially expressed genes between highly active immune cells and other cells were subsequently selected as features, and a least absolute shrinkage and selection operator (LASSO) regression model was constructed to screen for diagnostic-related features. Fourteen genes that were selected more than 5 times in 10 LASSO regression experiments were included in a multivariable Cox regression model. Finally, 3 genes (stathmin 1, cofilin 1, and C-C chemokine ligand 5) significantly associated with survival were identified and used to construct an immune-related gene signature. RESULTS: The immune-related gene signature composed of stathmin 1, cofilin 1, and C-C chemokine ligand 5 was identified through cell-cell communication. The effectiveness of the identified gene signature was validated based on experimental results of predictive immunotherapy response, tumor mutation burden analysis, immune cell infiltration analysis, survival analysis, and expression analysis. CONCLUSION: The findings suggest that the identified gene signature may contribute to a deeper understanding of the activity patterns of immune cells in the liver tumor microenvironment, providing insights for personalized treatment strategies.

ReHoGCNES-MDA: prediction of miRNA-disease associations using homogenous graph convolutional networks based on regular graph with random edge sampler.

Numerous investigations increasingly indicate the significance of microRNA (miRNA) in human diseases. Hence, unearthing associations between miRNA and diseases can contribute to precise diagnosis and efficacious remediation of medical conditions. The detection of miRNA-disease linkages via computational techniques utilizing biological information has emerged as a cost-effective and highly efficient approach. Here, we introduced a computational framework named ReHoGCNES, designed for prospective miRNA-disease association prediction (ReHoGCNES-MDA). This method constructs homogenous graph convolutional network with regular graph structure (ReHoGCN) encompassing disease similarity network, miRNA similarity network and known MDA network and then was tested on four experimental tasks. A random edge sampler strategy was utilized to expedite processes and diminish training complexity. Experimental results demonstrate that the proposed ReHoGCNES-MDA method outperforms both homogenous graph convolutional network and heterogeneous graph convolutional network with non-regular graph structure in all four tasks, which implicitly reveals steadily degree distribution of a graph does play an important role in enhancement of model performance. Besides, ReHoGCNES-MDA is superior to several machine learning algorithms and state-of-the-art methods on the MDA prediction. Furthermore, three case studies were conducted to further demonstrate the predictive ability of ReHoGCNES. Consequently, 93.3% (breast neoplasms), 90% (prostate neoplasms) and 93.3% (prostate neoplasms) of the top 30 forecasted miRNAs were validated by public databases. Hence, ReHoGCNES-MDA might serve as a dependable and beneficial model for predicting possible MDAs.

Fully Bio-Based 2,5-Furandicarboxylic Acid Polyester toward Plastics with Mechanically Robust, Excellent Gas Barrier and Fast Degradation.

Aliphatic-aromatic copolyesters offer a promising solution to mitigate plastic pollution, but high content of aliphatic units (>40 %) often suffer from diminished comprehensive performances. Poly(butylene oxalate-co-furandicarboxylate) (PBOF) copolyesters were synthesized by precisely controlling the oxalic acid content from 10 % to 60 %. Compared with commercial PBAT, the barrier properties of PBOF for H2O and O2 increased by more than 6 and 26 times, respectively. The introduction of the oxalic acid units allowed the water contact angle to be reduced from 82.5° to 62.9°. Superior hydrophilicity gave PBOF an excellent degradation performance within a 35-day hydrolysis. Interestingly, PBO20F and PBO30F also displayed obvious decrease of molecular weight during hydrolysis, with elastic modulus >1 GPa and tensile strength between 35-54 MPa. PBOF achieved the highest hydrolysis rates among the reported PBF-based copolyesters. The hydrolytic mechanism was further explored based on Fukui function analysis and density functional theory (DFT) calculation. Noncovalent analysis indicated that the water molecules formed hydrogen bonding interaction with adjacent ester groups and thus improved the reactivity of carbonyl carbon. PBOF not only meet the requirements of the high-performance packaging market but can quickly degrade after the end of their usage cycles, providing a new choice for green and environmental protection.

STGIC: A graph and image convolution-based method for spatial transcriptomic clustering.

Spatial transcriptomic (ST) clustering employs spatial and transcription information to group spots spatially coherent and transcriptionally similar together into the same spatial domain. Graph convolution network (GCN) and graph attention network (GAT), fed with spatial coordinates derived adjacency and transcription profile derived feature matrix are often used to solve the problem. Our proposed method STGIC (spatial transcriptomic clustering with graph and image convolution) is designed for techniques with regular lattices on chips. It utilizes an adaptive graph convolution (AGC) to get high quality pseudo-labels and then resorts to dilated convolution framework (DCF) for virtual image converted from gene expression information and spatial coordinates of spots. The dilation rates and kernel sizes are set appropriately and updating of weight values in the kernels is made to be subject to the spatial distance from the position of corresponding elements to kernel centers so that feature extraction of each spot is better guided by spatial distance to neighbor spots. Self-supervision realized by Kullback-Leibler (KL) divergence, spatial continuity loss and cross entropy calculated among spots with high confidence pseudo-labels make up the training objective of DCF. STGIC attains state-of-the-art (SOTA) clustering performance on the benchmark dataset of 10x Visium human dorsolateral prefrontal cortex (DLPFC). Besides, it's capable of depicting fine structures of other tissues from other species as well as guiding the identification of marker genes. Also, STGIC is expandable to Stereo-seq data with high spatial resolution.

Structural vaccinology, molecular simulation and immune simulation approaches to design multi-epitopes vaccine against John Cunningham virus.

The JCV (John Cunningham Virus) is known to cause progressive multifocal leukoencephalopathy, a condition that results in the formation of tumors. Symptoms of this condition such as sensory defects, cognitive dysfunction, muscle weakness, homonosapobia, difficulties with coordination, and aphasia. To date, there is no specific and effective treatment to completely cure or prevent John Cunningham polyomavirus infections. Since the best way to control the disease is vaccination. In this study, the immunoinformatic tools were used to predict the high immunogenic and non-allergenic B cells, helper T cells (HTL), and cytotoxic T cells (CTL) epitopes from capsid, major capsid, and T antigen proteins of JC virus to design the highly efficient subunit vaccines. The specific immunogenic linkers were used to link together the predicted epitopes and subjected to 3D modeling by using the Robetta server. MD simulation was used to confirm that the newly constructed vaccines are stable and properly fold. Additionally, the molecular docking approach revealed that the vaccines have a strong binding affinity with human TLR-7. The codon adaptation index (CAI) and GC content values verified that the constructed vaccines would be highly expressed in E. coli pET28a (+) plasmid. The immune simulation analysis indicated that the human immune system would have a strong response to the vaccines, with a high titer of IgM and IgG antibodies being produced. In conclusion, this study will provide a pre-clinical concept to construct an effective, highly antigenic, non-allergenic, and thermostable vaccine to combat the infection of the John Cunningham virus.

The Mathematical Modeling of the Host-Virus Interaction in Dengue Virus Infection: A Quantitative Study.

Infectious diseases, such as Dengue fever, pose a significant public health threat. Developing a reliable mathematical model plays a crucial role in quantitatively elucidating the kinetic characteristics of antibody-virus interactions. By integrating previous models and incorporating the antibody dynamic theory, we have constructed a novel and robust model that can accurately simulate the dynamics of antibodies and viruses based on a comprehensive understanding of immunology principles. It explicitly formulates the viral clearance effect of antibodies, along with the positive feedback stimulation of virus-antibody complexes on antibody regeneration. In addition to providing quantitative insights into the dynamics of antibodies and viruses, the model exhibits a high degree of accuracy in capturing the kinetics of viruses and antibodies in Dengue fever patients. This model offers a valuable solution to modeling the differences between primary and secondary Dengue infections concerning IgM/IgG antibodies. Furthermore, it demonstrates that a faster removal rate of antibody-virus complexes might lead to a higher peak viral loading and worse clinical symptom. Moreover, it provides a reasonable explanation for the antibody-dependent enhancement of heterogeneous Dengue infections. Ultimately, this model serves as a foundation for constructing an optimal mathematical model to combat various infectious diseases in the future.

Bioinformatic analysis of defective viral genomes in SARS-CoV-2 and its impact on population infection characteristics.

DVGs (Defective Viral Genomes) are prevalent in RNA virus infections. In this investigation, we conducted an analysis of high-throughput sequencing data and observed widespread presence of DVGs in SARS-CoV-2. Comparative analysis between SARS-CoV-2 and diverse DNA viruses revealed heightened susceptibility to damage and increased sequencing sample heterogeneity within the SARS-CoV-2 genome. Whole-genome sequencing depth variability analysis exhibited a higher coefficient of variation for SARS-CoV-2, while DVG analysis indicated a significant proportion of recombination sites, signifying notable genome heterogeneity and suggesting that a large proportion of assembled virus particles contain incomplete RNA sequences. Moreover, our investigation explored the sequencing depth and DVG content differences among various strains. Our findings revealed that as the virus evolves, there is a notable increase in the proportion of intact genomes within virus particles, as evidenced by third-generation sequencing data. Specifically, the proportion of intact genome in the Omicron strain surpassed that of the Delta and Alpha strains. This observation effectively elucidates the heightened infectiousness of the Omicron strain compared to the Delta and Alpha strains. We also postulate that this improvement in completeness stems from enhanced virus assembly capacity, as the Omicron strain can promptly facilitate the binding of RNA and capsid protein, thereby reducing the exposure time of vulnerable virus RNA in the host environment and significantly mitigating its degradation. Finally, employing mathematical modeling, we simulated the impact of DVG effects under varying environmental factors on infection characteristics and population evolution. Our findings provide an explanation for the close association between symptom severity and the extent of virus invasion, as well as the substantial disparity in population infection characteristics caused by the same strain under distinct environmental conditions. This study presents a novel approach for future virus research and vaccine development.

Immunoinformatics-driven In silico vaccine design for Nipah virus (NPV): Integrating machine learning and computational epitope prediction.

The Nipah virus (NPV) is a highly lethal virus, known for its significant fatality rate. The virus initially originated in Malaysia in 1998 and later led to outbreaks in nearby countries such as Bangladesh, Singapore, and India. Currently, there are no specific vaccines available for this virus. The current work employed the reverse vaccinology method to conduct a comprehensive analysis of the entire proteome of the NPV virus. The aim was to identify and choose the most promising antigenic proteins that could serve as potential candidates for vaccine development. We have also designed B and T cell epitopes-based vaccine candidate using immunoinformatics approach. We have identified a total of 5 novel Cytotoxic T Lymphocytes (CTL), 5 Helper T Lymphocytes (HTL), and 6 linear B-cell potential antigenic epitopes which are novel and can be used for further vaccine development against Nipah virus. Then we performed the physicochemical properties, antigenic, immunogenic and allergenicity prediction of the designed vaccine candidate against NPV. Further, Computational analysis indicated that these epitopes possessed highly antigenic properties and were capable of interacting with immune receptors. The designed vaccine were then docked with the human immune receptors, namely TLR-2 and TLR-4 showed robust interaction with the immune receptor. Molecular dynamics simulations demonstrated robust binding and good dynamics. After numerous dosages at varied intervals, computational immune response modeling showed that the immunogenic construct might elicit a significant immune response. In conclusion, the immunogenic construct shows promise in providing protection against NPV, However, further experimental validation is required before moving to clinical trials.

Improving the substrate binding of acetyl-CoA carboxylase (AccB) from Streptomyces antibioticus through computational enzyme engineering.

Malonyl-CoA serves as the main building block for the biosynthesis of many important polyketides, as well as fatty acid-derived compounds, such as biofuel. Escherichia coli, Corynebacterium gultamicum, and Saccharomyces cerevisiae have recently been engineered for the biosynthesis of such compounds. However, the developed processes and strains often have insufficient productivity. In the current study, we used enzyme-engineering approach to improve the binding of acetyl-CoA with ACC. We generated different mutations, and the impact was calculated, which reported that three mutations, that is, S343A, T347W, and S350W, significantly improve the substrate binding. Molecular docking investigation revealed an altered binding network compared to the wild type. In mutants, additional interactions stabilize the binding of the inner tail of acetyl-CoA. Using molecular simulation, the stability, compactness, hydrogen bonding, and protein motions were estimated, revealing different dynamic properties owned by the mutants only but not by the wild type. The findings were further validated by using the binding-free energy (BFE) method, which revealed these mutations as favorable substitutions. The total BFE was reported to be -52.66 ± 0.11 kcal/mol for the wild type, -55.87 ± 0.16 kcal/mol for the S343A mutant, -60.52 ± 0.25 kcal/mol for T347W mutant, and -59.64 ± 0.25 kcal/mol for the S350W mutant. This shows that the binding of the substrate is increased due to the induced mutations and strongly corroborates with the docking results. In sum, this study provides information regarding the essential hotspot residues for the substrate binding and can be used for application in industrial processes.

Structural and molecular investigation of the impact of S30L and D88N substitutions in G9R protein on coupling with E4R from Monkeypox virus (MPXV).

Understanding the pathogenesis mechanism of the Monkeypox virus (MPXV) is essential to guide therapeutic development against the Monkeypox virus. In the current study, we investigated the impact of the only two reported substitutions, S30L, D88N, and S30L-D88N on the G9R of the replication complex in 2022 with E4R using structural modeling, simulation, and free energy calculation methods. From the molecular docking and dissociation constant (KD) results, it was observed that the binding affinity did not increase in the mutants, but the interaction paradigm was altered by these substitutions. Molecular simulation data revealed that these mutations are responsible for destabilization, changes in protein packing, and internal residue fluctuations, which can cause functional variance. Additionally, hydrogen bonding analysis revealed that the estimated number of hydrogen bonds are almost equal among the wild-type G9R and each mutant. The total binding free energy for the wild-type G9R with E4R was -85.00 kcal/mol while for the mutants the TBE was -42.75 kcal/mol, -43.68 kcal/mol, and -48.65 kcal/mol respectively. This shows that there is no direct impact of these two reported mutations on the binding with E4R, or it may affect the whole replication complex or any other mechanism involved in pathogenesis. To explore these variations further, we conducted PCA and FEL analyses. Based on our findings, we speculate that within the context of interaction with E4R, the mutations in the G9R protein might be benign, potentially leading to functional diversity associated with other proteins.Communicated by Ramaswamy H. Sarma.

Structural basis for the mechanism of interaction of SARS-CoV-2 B.1.640.2 variant RBD with the host receptors hACE2 and GRP78.

The inflicted chaos instigated by the SARS-CoV-2 (Severe acute respiratory syndrome coronavirus 2) globally continues with the emergence of novel variants. The current global outbreak is aggravated by the manifestation of novel variants, which affect the effectiveness of the vaccine, attachment with hACE2 (human Angiotensin-converting enzyme 2) and immune evasion. Recently, a new variant named University Hospital Institute (IHU) (B.1.640.2) was reported in France in November 2021 and is spreading globally affecting public healthcare. The B.1.640.2 SARS-CoV-2 strain revealed 14 mutations and 9 deletions in spike protein. Thus, it is important to understand how these variations in the spike protein impact the communication with the host. A protein coupling approach along with molecular simulation protocols was used to interpret the variation in the binding of the wild type (WT) and B.1.640.2 variant with hACE2 and Glucose-regulating protein 78 (GRP78) receptors. The initial docking scores revealed a stronger binding of the B.1.640.2-RBD with both the hACE2 and GRP78. To further understand the crucial dynamic changes, we looked at the structural and dynamic characteristics and also explored the variations in the bonding networks between the WT and B.1.640.2-RBD (receptor-binding domain) in association with hACE2 and GRP78, respectively. Our findings revealed that the variant complex demonstrated distinct dynamic properties in contrast to the wild type due to the acquired mutations. Finally, to provide conclusive evidence on the higher binding by the B.1.640.2 variant the TBE was computed for each complex. For the WT with hACE2 the TBE was quantified to be-61.38 ± 0.96 kcal/mol and for B.1.640.2 variant the TBE was estimated to be -70.47 ± 1.00 kcal/mol. For the WT-RBD-GRP78 the TBE -was computed to be 32.32 ± 0.56 kcal/mol and for the B.1.640.2-RBD a TBE of -50.39 ± 0.88 kcal/mol was reported. This show that these mutations are the basis for higher binding and infectivity produced by B.1.640.2 variant and can be targeted for drug designing against it.Communicated by Ramaswamy H. Sarma.

De novo generation of dual-target ligands for the treatment of SARS-CoV-2 using deep learning, virtual screening, and molecular dynamic simulations.

De novo generation of molecules with the necessary features offers a promising opportunity for artificial intelligence, such as deep generative approaches. However, creating novel compounds having biological activities toward two distinct targets continues to be a very challenging task. In this study, we develop a unique computational framework for the de novo synthesis of bioactive compounds directed at two predetermined therapeutic targets. This framework is referred to as the dual-target ligand generative network. Our approach uses a stochastic policy to explore chemical spaces called a sequence-based simple molecular input line entry system (SMILES) generator. The steps in the high-level workflow would be to gather and prepare the training data for both targets' molecules, build a neural network model and train it to make molecules, create new molecules using generative AI, and then virtually screen the newly validated molecules against the SARS-CoV-2 PLpro and 3CLpro drug targets. Results shows that novel molecules generated have higher binding affinity with both targets than the conventional drug i.e. Remdesivir being used for the treatment of SARS-CoV-2.Communicated by Ramaswamy H. Sarma.

TEPCAM: Prediction of T-cell receptor-epitope binding specificity via interpretable deep learning.

The recognition of T-cell receptor (TCR) on the surface of T cell to specific epitope presented by the major histocompatibility complex is the key to trigger the immune response. Identifying the binding rules of TCR-epitope pair is crucial for developing immunotherapies, including neoantigen vaccine and drugs. Accurate prediction of TCR-epitope binding specificity via deep learning remains challenging, especially in test cases which are unseen in the training set. Here, we propose TEPCAM (TCR-EPitope identification based on Cross-Attention and Multi-channel convolution), a deep learning model that incorporates self-attention, cross-attention mechanism, and multi-channel convolution to improve the generalizability and enhance the model interpretability. Experimental results demonstrate that our model outperformed several state-of-the-art models on two challenging tasks including a strictly split dataset and an external dataset. Furthermore, the model can learn some interaction patterns between TCR and epitope by extracting the interpretable matrix from cross-attention layer and mapping them to the three-dimensional structures. The source code and data are freely available at https://github.com/Chenjw99/TEPCAM.

Omicron variant (B.1.1.529) challenge the integrity of blood brain barrier: Evidence from protein structural analysis.

Studies on nonhuman primates, wild-type and transgenic mice have shown the presence of SARS-CoV-2 RNA components in the brains. Despite the Blood-Brain Barrier (BBB) provides protection there are less evidences on how the SARS-CoV-2 crosses the BBB. Given that there is an increase of Omicron reinfection rates, transmissibility rate and involvement to cause neurological dysfunctions, we hypothesized to investigate how the Omicron variant (B.1.1.529) binds structurally to key BBB-maintaining proteins and thus can possibly challenge the integrity and transportation to the brain. By using molecular dynamics simulation approaches we examined the interaction of Omicron variant (B.1.1.529) with different structural and transporter proteins located at the BBB. Our results show that in Zona Ocludin 1-RBD complex, we observe a distinct pattern. Omicron demonstrates a docking score of -88.9 ± 6.8 kcal/mol and six interactions, while the wild type (WT) presents a higher score of -94.0 ± 2.3 kcal/mol, forming eight interactions. Comparing affinities, the WT-RBD displays a stronger preference for Claudin-5, boasting a docking score of -110.2 ± 3.0 and nine interactions, versus Omicron-RBD's slightly reduced engagement, with a docking score of -105.6 ± 0.2 and seven interactions. Interestingly, the Omicron variant exhibits heightened stability in interactions with Glucose Transporter and ABC transporters, registering docking scores of -110.6 ± 1.9 and -112.0 ± 3.6 kcal/mol, respectively. This surpasses the WT's respective scores of -95.2 ± 2.2 and -104.0 ± 6.2 kcal/mol, reflecting a unique interaction profile. Rigorous molecular dynamics simulations validate our findings. Our study emphasizes the Omicron variant's increased affinity towards transporter proteins, illuminating potential implications for BBB integrity and brain transportation. While these insights offer a valuable framework, comprehensive experimental validation is indispensable for a comprehensive understanding.

A Multimodal Deep Learning Framework for Predicting PPI-Modulator Interactions.

Protein-protein interactions (PPIs) are essential for various biological processes and diseases. However, most existing computational methods for identifying PPI modulators require either target structure or reference modulators, which restricts their applicability to novel PPI targets. To address this challenge, we propose MultiPPIMI, a sequence-based deep learning framework that predicts the interaction between any given PPI target and modulator. MultiPPIMI integrates multimodal representations of PPI targets and modulators and uses a bilinear attention network to capture intermolecular interactions. Experimental results on our curated benchmark data set show that MultiPPIMI achieves an average AUROC of 0.837 in three cold-start scenarios and an AUROC of 0.994 in the random-split scenario. Furthermore, the case study shows that MultiPPIMI can assist molecular docking simulations in screening inhibitors of Keap1/Nrf2 PPI interactions. We believe that the proposed method provides a promising way to screen PPI-targeted modulators.