The primary sequences of the G, F, M, L, P, and N proteins of NiV were obtained from the NCBI database (accessed on October 30, 2023) using the following accession IDs: NP_112027.1, NP_112026.1, NP_112025.1, NP_112028.1, NP_112022.1, and NP_112021.1. Their antigenicity was assessed using VaxiJen v.2.0 (accessed on October 30, 2023) (37). A threshold score of 0.4 was established as the cutoff for viral proteins, where proteins scoring higher than 0.4 were classified as antigenic, while those scoring less than 0.4 were deemed non-antigenic. Apart from the above protein sequences, the Cholera toxin subunit B and Beta-defensin 3 adjuvant protein sequences were obtained from the UniProt database, with the respective UniProt entry identifiers P01556 and Q5U7J2.

The IEDB server (https://www.iedb.org/) was utilized for the identification of MHC-I/CTL and MHC-II/HTL epitopes within the six structural proteins of NiV (G, F, M, N, L, and P) (38). Cytotoxic T lymphocyte (CTL) and helper T lymphocyte (HTL) epitope predictions were conducted utilizing the NetMHCpan 4.0 EL and NetMHCIIpan 4.1 EL methods, respectively (39). For CTL epitopes, a peptide length ranging from 9 to 10 was taken into consideration, whereas for HTL epitopes, a length of 15 was utilized, both in conjunction with the reference set of alleles. The reference set of HLA alleles, signifying commonly shared binding specificities, was implemented to ensure population coverage exceeding 97% and 99% for MHC class I and II, respectively. Additionally, a rigorous criterion was established, with a percentile rank of less than 1 for MHC class I and less than 10 for MHC class II alleles, denoting robust binding peptides during the epitope screening. To validate the selected epitopes for their antigenicity, VaxiJen v.2.0 was employed, employing the default prediction value of 0.4 (37).

Moreover, the MHC I Immunogenicity Tool accessible at IEDB (http://tools.iedb.org/immunogenicity/) was utilized to project the immunogenicity of CTL epitopes (40). For further scrutiny, only CTL epitopes with positive immunogenicity score were taken into consideration. Conversely, HTL epitopes were analyzed for their potential to elicit IFN-γ cytokine production, contributing to the initiation of innate and adaptive immunity against the virus through the employment of the IFNepitope server (IFNepitope: A server for predicting and designing IFN-gamma inducing epitopes (osdd.net) (41). Epitopes demonstrating a “positive” response in terms of IFNγ release were exclusively selected for subsequent analysis.

To identify potential B cell epitopes within the NiV proteins, we utilized the BCPREDS server 1.1 (http://ailab-projects1.ist.psu.edu:8080/bcpred/predict.html), a web-based B cell epitope prediction tool (42). Epitope prediction focused on 16-mer peptides, with a selection threshold of prediction scores exceeding 0.90. The predicted epitopes were further subjected to antigenicity evaluation using the VaxiJen v2.0 server (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html) (37).

The identified B-cell, CTL, and HTL epitopes underwent further analysis to refine the selection based on their conservation, allergenicity, and toxicity. To assess conservation levels, the IEDB (Epitope Conservancy Analysis at iedb.org) was employed for the conservation analysis of B and T-cell epitopes against the available sequences of both NiV strain B and M for each structural protein of NiV (G, F, M, N, L) (43). Only epitopes demonstrating 100% conservation were considered for subsequent property analysis.

AllerTop 2.0, accessible at http://www.pharmfac.net/allertop, was utilized to evaluate the allergenicity of B and T-cell epitopes. AllerTOP v2.0 utilizes auto- and cross-covariance (ACC) transformation, amino acid E-descriptors, and k nearest neighbor machine learning techniques for protein allergen classification. Epitopes identified as “probable non-allergenic” were specifically chosen for our analysis (44). Additionally, the ToxinPred online server, available at https://webs.iiitd.edu.in/raghava/toxinpred/index.html, was employed to predict the toxicity of B and T-cell epitopes (45).

The distribution of particular HLA alleles among various ethnicities and populations is crucial for developing an epitope-based vaccine. The IEDB population coverage tool (http://tools.iedb.org/population/) was utilized to assess the population coverage of the selected CTL and HTL epitopes across multiple HLA alleles in diverse global regions as well as on a worldwide level, calculating combined class I and II coverage (46).

The proposed design of the vaccine construct involved connecting an adjuvant to specific T and B-cell epitopes, which are interlinked by specific linkers to accurately define the epitopes. Various linkers utilized in this study, such as CTL linker (AAY), B epitope linker (KK), and HTL linker (GPGPG), are chosen to achieve solubility, improved expression levels, and precise folding of the multi-epitope structure (47, 48). Each multi-epitope vaccine was integrated with a potent immunostimulatory adjuvant to boost immunogenicity and activate both innate and adaptive immune responses (49). Two different adjuvants were employed in this study: CTB (Accession ID: P01556) and beta defensin (Accession ID: Q5U7J2). The adjuvant was linked to the N-terminal of the vaccine construct using the EAAAK linker (50). The final structure of the vaccine constructs consist of the adjuvant, along with CTL, B-cell, and HTL epitopes, arranged in that specified sequence. Following the assembly of the vaccine constructs, antigenicity was evaluated using VaxiJen v 2.0 (37), while their allergenicity was assessed using AllerTop 2.0 (44). Ultimately, the toxicity of the formulated vaccines were predicted using the ToxinPred web server (45).

ExPASy’s ProtParam tool (https://web.expasy.org/protparam/) was used to analyze the physical and chemical characteristics of vaccine construct (51). A number of parameters were considered, including aliphatic index, stability index, Grand Average of Hydropathicity (GRAVY), and half-life. The instability index was calculated to determine the stability of epitopes in vivo, with a threshold of <40 indicating a stable vaccine (52). Additionally, SoluProt 1.0 (https://loschmidt.chemi.muni.cz/soluprot/) was used to evaluate the solubility of the vaccine (53).

The PSIPRED software (http://bioinf.cs.ucl.ac.uk/psipred/) was employed to predict the structures of alpha helix, beta helix, and coil in the vaccine constructs (54). The tertiary structure of the multi-target, multi-epitope vaccine peptide was generated using Robetta server (55), utilizing the RoseTTAFold algorithm for de novo protein modeling (https://robetta.bakerlab.org). RoseTTAfold is a tool that employs neural networks and simultaneously takes into account the arrangement of amino acids, their interactions, and the potential tertiary structures they can form (56).

To enhance the quality of the 3D modeled structure, a refinement process was applied using the GalaxyRefine online server (http://galaxy.seoklab.org/refine) (57). This process involves rebuilding and repacking of the side chains, followed by structure relaxation through molecular dynamics (MD) simulation. Validating the 3D structure after refinement provides an overview of the structure improvement and its current quality. Tools such as PROCHECK, the ERRAT tool from the UCLA-DOE LAB (https://saves.mbi.ucla.edu/) (58, 59), and the ProSA-web server (https://prosa.services.came.sbg.ac.at/prosa.php) were utilized to assess the validity and quality of the selected 3D structure (60).

To identify potential epitopes capable of inducing B-cell production, the conformational B-cell epitopes were scrutinized for the final vaccine construct following 3D structure prediction and refinement. For the prediction of these epitopes, the ElliPro: Antibody Epitope Prediction tool from the IEDB database (tools.iedb.org/ellipro/) was utilized, leveraging protein structures or sequences to identify discontinuous antibody epitopes (61).

For analyzing the binding pattern of the multi-epitope vaccine polypeptide with receptors TLR-3, molecular docking analysis was carried out using the ClusPro 2.0 server (cluspro.bu.edu/login.php) (62). To facilitate this, TLR-3 (PDB ID: 1ziw) was obtained from RCSB PDB (RCSB PDB: Homepage) (63). ClusPro 2.0, an online server, employs a combination of shape-based and energy-based algorithms to predict the binding affinity of a protein-ligand complex. It analyzes the 3D structure of proteins, predicts binding sites, and identifies potential interactions between the proteins and ligands. Furthermore, to predict interacting residues involved in the molecular interactions of the designed vaccine construct and receptors, the online database PDBsum was utilized (64).

Examining the binding interactions between the vaccine and the receptor TLR3 requires considering their dynamic behavior. Consequently, MD simulations were conducted using the Desmond module from the Schrodinger suite (65, 66). The vaccine-TLR3 complex underwent simulation in a TIP3P predefined water solvent model within orthorhombic periodic boundary conditions. To achieve electrical neutralization, the addition of appropriate amounts of sodium and chlorine ions was performed. The system was then subjected to minimization under the OPLS4 force field (67). The NPT ensemble (68) was employed for the simulation, maintaining a constant temperature of 300 K at 1 atm pressure over a duration of 100 ns (69).

VectorBuilder Codon optimization tool, available at https://en.vectorbuilder.com/tool/codon-optimization, was utilized to evaluate the expression level of the multi-epitope vaccine in E. coli (strain K12). VectorBuilder conducted analyses for the GC content and Codon Adaptation Index value of the query sequence with the goal of achieving optimal expression. Subsequently, the final vaccine constructs were inserted into the pET-28a(+) plasmid using SnapGene software v5.2.3 (https://www.snapgene.com/).

To assess the capability of the designed vaccine to induce a sustained immune response, C-IMMSIM analyses were conducted using the webserver accessible at https://kraken.iac.rm.cnr.it/CIMMSIM/index.php?page=1. The analysis included the calculation of an immunogenic response following a single-dose injection. C-ImmSim employs an agent-based model incorporating machine learning techniques and immune epitope prediction. It utilizes a position-specific scoring matrix (PSSM) to enable the prediction of immune interactions (70).

Antigenicity is a crucial parameter that plays a significant role in the evaluation of proteins for vaccine development. Based on antigenicity screening of the complete amino acid sequences of the six NiV structural proteins (Glycoprotein, Fusion, Matrix, Nucleocapsid, Phosphoprotein, and Polymerase) using VaxiJen v.2.0 (37), all structural proteins (G, F, M, N, P, L) were projected to be antigenic with scores exceeding the 0.4 threshold, as demonstrated in Table 1 . Consequently, these proteins were chosen for B- and T-cell epitope prediction and vaccine development.

The aim of T-cell epitope prediction is to identify the shortest peptide sequences within an antigen that can activate either CD4 or CD8 T-cells (71). Antigen-presenting cells display T-cell epitopes on their surface where they attach to MHC class I or MHC-II proteins. T-cell epitopes that bind to MHC class I proteins are acknowledged by CD8 T-cells which differentiate into cytotoxic T lymphocytes (CTL), whereas those presented by MHC class II are recognized by CD4 T-cells, which are identified as helper T-cells (72).

The IEDB server was employed to analyze the antigenic sequences of all structural proteins using the NetMHCpan 4.0 EL and NetMHCIIpan 4.1 EL methods for identifying T-cell epitopes (CTL and HTL), with the HLA reference set. The selection of the complete HLA reference set aimed to generate epitopes that cover a broad range of globally prevalent MHC class I and II alleles. Based on the selected threshold for CTL and HTL, NetMHCpan 4.0 EL predicted 778 epitopes for G protein, 756 for F, 453 for M, 713 for N, 665 for P, 4569 for L protein. Similarly, NetMHCIIpan 4.1 EL used for HTL epitope prediction, identified 1312 epitopes for G, 1273 for F, 840 for M, 1290 for N, 1650 for P, and 5831 for L. Subsequent to these predictions, the most appropriate CTL and HTL epitopes were selected based on criteria including antigenicity, non-toxicity, complete conservation across all strains, and absence of allergenic properties, while also taking into account the frequency of interactions with various alleles. The immunogenic potential and ability to elicit cytotoxic T-cell responses were evaluated, employing the MHC class I Immunogenicity tool available in the IEDB server for CTL-specific epitopes. The IFNepitope server was utilized to examine the capacity to stimulate IFN-gamma production for HTL-specific epitopes. In consideration of these parameters, the top two epitopes for each protein were selected. A total 12 CTL and 12 HTL epitopes were considered for the designing of the vaccine construct, as detailed in Tables 2 , 3 .

The role of B-cell epitopes is vital in the multi-epitope vaccine as they trigger B-lymphocytes to generate antibodies, a crucial aspect of adaptive immunity. To predict linear B-cell epitopes from all structural proteins, the BCPREDS tool was utilized. Following prediction, the identified epitopes were rigorously assessed for antigenicity, non-allergenicity, and non-toxicity. Epitopes that met these criteria and exhibited high antigenicity scores were considered suitable. A total of six B-cell epitopes, as detailed in Table 4 , were shortlisted from the six structural proteins for inclusion in the vaccine construct. Every chosen epitope from each protein exhibited complete identity with all retrieved amino acid sequences of the respective protein. Therefore, the shortlisted epitopes were predicted to possess a cross-reactivity against both NiV M and B strains.

The distribution of MHC HLA alleles varies across different geographic regions and ethnic populations worldwide. Thus, it is important to consider population coverage when designing a vaccine for maximum effectiveness. In this study, totally 24 (12 CTL and 12 HTL) identified epitopes were analyzed for population coverage at the worldwide level using IEDB population coverage tool. Epitopes identified to bind to multiple MHC alleles are deemed optimal only if their combined frequency within a population demonstrates substantial coverage, nearing 100% or achieving close to it. This is important for ensuring that a vaccine targeting those epitopes would be effective in a majority of individuals within the population.

The shortlisted CTL and HTL epitopes covered 98.42%and 99.68% of the global population, respectively. Overall, the significant population coverage was observed for the chosen 24 epitopes. These epitopes were found to cover 99.99% of resultant alleles of the world population ( Supplementary Figure 1 ). The highest coverage of 100% was identified across five distinct regions worldwide, spanning East and West Africa, Europe, and North and South America, as determined by the combined CTL and HTL epitopes analysis. Additionally, during population coverage prediction, we noted the prevalence in regions highly impacted with previous Nipah outbreaks: South, Southeast Asia. The population data for these regions encompassed the global population, indicating potential extensive coverage for countries like India, Bangladesh, Malaysia, Singapore, and the Philippines, where Nipah outbreaks are common. South Asia indicated a coverage rate of 99.99%, with strong CTL and HTL coverage rates of 94.91% and 99.74% respectively. Conversely, Southeast Asia also demonstrated a strong coverage rate of 99.70% for combined CTL and HTL epitopes ( Figure 2 ).

Two distinct constructs were designed by integrating antigenic epitopes with two different adjuvants (Cholera Toxin B, and Beta-defensin) and linkers. While CTB and Beta-defensin 3 may pose potential risks to vaccine efficacy and immune response, these can be mitigated through optimized dosing and controlled delivery (73, 74). In this study, CTB was chosen over whole Cholera Toxin due to its non-toxic nature and its ability to enhance immune activation (75). Given the impracticality of evaluating numerous potential epitope arrangements, we selected representative constructs for the analysis. The sequences of these constructs differed according to the adjuvant employed and the arrangement of the epitopes. These constructs, comprising 12 CTL, 6 B-cell, and 12 HTL epitopes, were interconnected using universal linkers (EAAAK, AAY, KK, and GPGPG) as represented in Figure 3 (47, 48).

An ideal vaccine should exhibit antigenicity, be non-allergenic, and free from toxins. Both vaccine candidates (NiV_1-2) were predicted to be antigenic by the VaxiJen v2.0 server. Furthermore, analysis through the AllerTOP v2.0 and ToxinPred servers indicated that the constructs are likely non-allergenic and non-toxic, respectively.

The physicochemical properties were assessed using various parameters, including amino acid composition, molecular weight, theoretical pI, estimated half-life, instability index (I.I), aliphatic index (A.I), and the grand average of hydropathicity (GRAVY), as illustrated in Table 5 . The lengths of the vaccine constructs NiV_1 and NiV_2 consist of 623 and 544 amino acids, respectively, with predicted molecular weights ranging from 59.37 to 68.35 kDa. Both constructs are classified as basic, with theoretical pI values between 9.50 and 9.69. Among the two constucts, NiV_2 had the highest predicted solubility (0.82) and theoretical pI (9.69), as well as elevated antigenicity. For both NiV_1 and NiV_2 constructs, the estimated half-life in mammalian reticulocytes, representing an in vitro environment, is approximately 30 hours, while in yeast it is predicted to exceed 20 hours and in Escherichia coli over 10 hours, simulating in vivo conditions. The instability index is calculated, indicating protein stability. The GRAVY score ranges from -0.321 to -0.407, reflecting the hydrophilic nature of the vaccine constructs, suggesting favorable interaction with surrounding water molecules. In summary, the additional physicochemical parameters of the proposed multi-epitope vaccine candidates fall within acceptable ranges, reinforcing their potential as viable vaccine candidates.

The stability of the vaccine candidates in real-world conditions was evaluated by using the 623 and 544 amino acid peptide sequences to predict both the secondary and tertiary structures. According to results from the PSIPRED server, the secondary structure of the designed vaccine constructs comprises numerous helices, a small number of strands, and a significant portion of coils ( Supplementary Figures 2A, B ). The secondary structure analysis of NiV_1 and NiV_2 revealed composition of 33.70% and 30.69% helices, 48.31% and 49.09% coils, and 17.97% and 20.22 strands, respectively.

The Robetta server, employing the RoseTTAFold algorithm, was utilized to predict the tertiary structure of the proposed vaccines, resulting in five potential 3D structures. The optimal model for the vaccines was selected based on criteria that included having the highest proportion of residues in the favorable regions of the Ramachandram plot and the lowest percentage in the outlier region, indicating superior structural quality. Model 1 and 3 of NiV_1 and NiV_2 emerged as the best among the RoseTTAFold-generated models. For NiV_1, the Ramachandran plots illustrating that 88.6% of residues lie in the most favored region, 9.1% in the allowed region, and only 1.2% in the disallowed region ( Supplementary Figure 3 ). To further enhance the stability of the predicted structure and increase the number of residues in the favorable region, the structure was subjected to refinement using the GalaxyRefine server. The refined structure was validated using the Ramachandran plot via the PROCHECK server, revealing notable improvements. In the best model of NiV_1 ( Figure 4A ), 91.3% of residues fell within the favored region, with 6.4% in the allowed region, and only 1.4% in the outlier region.

Conversely, the model of the NiV_2 exhibited 87.7%, 9.8%, 0.9%, 1.6% residues in the most favored, additionally allowed, generously allowed, and disallowed regions, respectively ( Supplementary Figure 3 ). The refined model of NiV_2 demonstrated 90.5%, 7.3%, 0.5%, 1.8% residues in the Ramachandran favored, additionally allowed, generously allowed, and disallowed regions, respectively ( Figure 4A ). Furthermore, ProSA-web analysis was conducted to assess quality and identify potential flaws in the top models. This analysis computes z-scores, which validate the modeled protein’s alignment with similarly sized natural proteins determined by NMR or X-ray methods. The initial models of the NiV_1 and NiV_2 yielded a z-score of -8.17 and -5.73, which changed to -8.27 and -5.78 post-refinement, respectively ( Figure 4B ). These z-scores fall within the acceptable range for protein structures of similar size. Moreover, the overall quality factor of the designed vaccine models, as determined by ERRAT analysis, were 94.155 and 97.276, indicating a high-quality model suitable for further analysis ( Figure 4C ), including molecular docking, and simulations. ProSA-web Z-score and Ramachandran plot details of NiV_1–2 are provided in Figure 4 and Table 6 .

The 3D structures of the proposed vaccine candidates, NiV_1 and NiV_2, were screened for conformational B-cell epitopes using the ElliPro webserver at 0.5 threshold, identifying 7 and 9 epitopes, respectively ( Supplementary Figures 4 , 5 ). The predicted scores for these epitopes ranged from 0.612 to 0.939 for NiV_1 ( Table 7 ) and from 0.507 to 0.88 for NiV_2 ( Table 8 ).

TLR3 is a member of the toll-like receptor family that recognizes double-stranded RNA and activates downstream signaling to induce antiviral immune responses (76). The Nipah virus has evolved a process to block the TLR3 signaling pathway, leading to a compromise in the host’s ability to defend against viruses (77). To overcome this evasion strategy, it is crucial to assess the ability of the designed vaccine to bind to the TLR-3 immune receptor via molecular docking analysis using ClusPro 2.0 webserver. Docking simulations generated 10 potential complexes for each vaccine construct, with varying energy scores. The final complexes for NiV_1 and NiV_2 were selected based on the lowest energy scores of -1284.8 kcal/mol and -1222.6 kcal/mol, respectively, indicating strong binding affinity between the designed vaccines and immune receptors. The molecular interactions between the proposed vaccines and the TLR receptor are depicted in Figure 5 . The NiV_1 complex exhibited robust interactions, including 4 salt bridges, 32 hydrogen bonds, and 343 non-covalent interactions. Similarly, the NiV_2 vaccine formed 4 salt bridges, 10 hydrogen bonds, and 182 non-covalent interactions with the TLR3 receptor. The detailed atom-level interactions between TLR3 and NiV_1 & 2 are provided in Supplementary Tables 1 , 2 .

To determine the structural stability and the dynamic behavior of the TLR3-vaccine docked complexes, molecular dynamic simulation was performed for 100ns using Desmond Schrodinger software. For TLR3-NiV_1 docked complex, Root mean square deviation (RMSD) analysis showed variation up to about 25 ns; thereafter the docked vaccine-TLR3 complex maintained stability with less than 1Å deviation. In contrast, the TLR3-NiV_2 complex exhibited fluctuations (RMSD ≥ 9Å) throughout the simulation, implying its instability compared to the TLR3-NiV_1 complex ( Figure 6A ). These findings align with the docking analysis, indicating stronger binding interactions between NiV_1 and TLR3. Additionally, Root-mean-square fluctuations (RMSF) of the complexes was also performed to identify the flexibility across the amino acid residues in the TLR3-vaccine complexes. A high degree of fluctuations were observed in both of the vaccine molecules compared to the TLR3 molecule ( Figure 6B ), which exhibited mostly rigidity and low RMSF values. The average RMSF of TLR3-NiV_1 complex was calculated as 2.62Å, with ASP354 (13.02 Å; vaccine residue) showing highest fluctuation among residues analyzed ( Figure 6B ). The TLR3-NiV_2 complex demonstrated flexibility throughout the simulation, with RMSF values ranging 2 to 6Å ( Figure 6B ). Overall, the molecular dynamics simulation suggests that the NiV_1 multiepitope vaccine construct demonstrates stronger and more stable interactions with the TLR3 immune receptor. The study also indicates that residues in the adjuvants, CTL, B-cell, and HTL epitopes show increased flexibility, which seems to be essential for the designed vaccine to adopt a suitable conformation for interacting with immune cells.

After analyzing both vaccine constructs, we finalized the NiV_1 vaccine construct based on the molecular docking and molecular dynamics simulation results. The designed vaccine construct’s reverse translation and codon optimization were conducted using the VectorBuilder Codon optimization tool. Utilizing the Codon optimization web tool, an assessment of significant gene sequence properties for achieving high-level protein expression in the E. coli host was performed, including estimation of Codon Adaptation Index (CAI) and GC content. The optimized codon sequence comprised 1872 nucleotides. A codon adaptation index value greater than 0.8 and a GC content ranging from 30% to 70% are considered conducive for effective protein expression in the host system. The analysis revealed that the designed vaccine construct obtained a CAI value of 0.94, with a desirable GC content of 53.85% that can help achieve high protein expression. Furthermore, restriction sites XhoI and NdeI were incorporated into both the N and C terminals of the final vaccine codon sequence before being inserted into the pET-28a(+) vector using the SnapGene tool ( Figure 7 ). The graphical representation in Figure 7 depicts the successful cloning of NiV_1 vaccine construct into the pET-28a(+) expression vector.

The C-ImmSim web server was utilized to forecast the immune response profile for the suggested vaccine construct. The findings indicate that the construct has the potential to trigger an immune reaction. The graph demonstrates that within 5 days of reaching approximately 700,000 antigens per mL, the antigen is eliminated from the host system. It is evident that during the first 15 days following injection when antigen count was close to zero, there was a rise in production rate of primary antibodies as well as IgM and IgG levels increased up to around 9000. Subsequently, although the complex count reduced, it stabilized at approximately 4500 after one-month indicating sustained and long-lasting immunity response ( Figure 8A ). Following vaccination, the total B cell count exceeded 450 cells/mm³, and memory B cell and IgM levels persisted throughout the simulation ( Figure 8B ). Plasma B cell levels reached a peak of over 7 cells/mm³ between days 5 and 10, with subsequent elevations in IgM and IgG1 levels. Notably, Figure 8C highlights a substantial increase in plasma B cell production, underscoring its critical role in eliminating pathogens.

The vaccine construct was also examined for its ability to induce cytokine production including interferon gamma, interleukins etc. Figure 8D , depicts the release of 400000 and 200000 ng\mL concentration of interferon gamma and interleukins, respectively, generated by the injection of the proposed vaccine candidate. Additionally, the populations of helper T (Th) and cytotoxic T (Tc) cells significantly increased post-vaccination, as depicted in Figures 8E, F . According to C-ImmSim simulations, the potential NiV_1 vaccine construct can induce a durable cellular and humoral immune response. Nevertheless, experimental validation is imperative to confirm its ability to elicit adaptive immunity against NiV_1.