{"paper_id":"1f774b51-6148-4ef7-97ca-b12f8df41bac","body_text":"1\nComputational Prediction of Plasmodium falciparum Antigen-T-cell Receptor \nInteractions via Molecular Docking: Implications for Malaria Vaccine Design\nGilbert Kipkoech1*, Wesley Kanda1, Beatrice Irungu1, Mary Nyangi1, Cecilia Kimani1, Ruth \nNyangacha1, Lucia Keter1, Diana Atieno1, Jeremiah Gathirwa1, Elizabeth Kigondu1, Edwin \nMurungi2\n1. Centre for Traditional Medicine and Drug Research, Kenya Medical Research \nInstitute\n2. Kisii University, Kenya\n*Corresponding Author: Kipkoech Gilbert (kipkoech.gillie@gmail.com) \nAbstract\nMalaria is one of the deadliest diseases in sub-Saharan Africa and Southeast Asia. The majority of \nthe fatalities occur mostly in children under 5 years and pregnant women and this is due to infection \nby Plasmodium spp, of which Plasmodium falciparum is the most virulent and is responsible for \nmost of the morbidity and mortality. Despite various public health interventions such as use of \ninsecticide-treated bed nets, spraying of homes with insecticides and use of WHO recommended \nartemisinin-based combination therapies (ACT), malaria prevention still faces major setback due \nto drug and insecticide resistance by P. falciparum and mosquitoes respectively. The study uses \nmolecular docking and immunoinformatics to screen various Plasmodium spp antigens and \nevaluate their antigenicity and suitability as vaccine candidates. The P. falciparum antigens and \nT-cell receptor (TCR) structures were obtained from Protein Data Bank (PDB) based on a range \nof factors related to their role in the lifecycle of the parasite and their status as vaccine targets. \nProtein structures not available in the PDB were predicted using AlphaFold. The 3D structures of \nselected P. falciparum antigens and TCR structures were downloaded in PDB format then all water \nmolecules, Hetatm, and bound ligands were deleted from the protein structures using BIOVIA \nDiscovery Studio Visualizer. Subsequently, molecular docking was done using ClusPro v2.0 \nserver and docked complexes were compared. The findings of this study gave valuable insights \ninto the interaction of human immune response with P. falciparum antigens. The best three ranked \nantigen complexes are PfCyRPA, PfMSP10 and PfCSP and this confirm their use as potential \ncandidates for vaccine development. This study highlights the usefulness of computational \ndocking in identifying P. falciparum antigens of excellent immunogenic potential as vaccine \ncandidates.\nKeywords: malaria, Plasmodium falciparum, artemisinin-based combination therapies, drug and \ninsecticide resistance, molecular docking, immunoinformatics, T-cell receptor (TCR), ClusPro \nv2.0, PfCyRPA, PfMSP10 and PfCSP.\nIntroduction\nBackground on malaria and Plasmodium falciparum\nThe global burden of malaria remains huge particularly in sub-Saharan Africa and Southeast Asia \nwhere majority of the fatalities occur mostly in children under 5 years and pregnant women [1]. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n2\nAccording to World Health Organization (WHO) report of 2022, it is estimated that there were \n249 million cases of malaria with 608,000 deaths reported in 85 countries. African nations are \nimmensely disproportionately affected and account for 94% of the total reported cases and 95% of \ndeaths [2]. The causative agent for malaria is Plasmodium spp, of which Plasmodium falciparum \nis the most virulent and is responsible for most of the morbidity and mortality. Infection \ntransmission is initiated when infected female anopheles mosquitoes bite non-infected humans \ntransferring sporozoites that undergo developmental stages in humans to become infective. \nAnopheles gambiae is the dominant transmission vector in most areas [3]. Despite an array of \npublic health interventions aimed at halting transmission such as insecticide-treated bed nets, \nspraying of homes with insecticides and use of WHO recommended artemisinin-based \ncombination therapies (ACT), malaria prevention faces the main setback drug and insecticide \nresistance by P. falciparum and mosquitoes respectively. Thus, the development of novel malaria \ninterventions is a pressing priority.\nThe life cycle of Plasmodium falciparum is distinctly complex, comprising of numerous stages in \nthe mosquito and human hosts [4]. When entering the human host, P. falciparum infects red blood \ncells, which are killed and thus start causing symptoms of severe disease, including anemia, \ncerebral malaria, and multi-organ failure [5,6]. It is worth noting that besides being resistant to \nmajority of the traditional antimalarial agents, the pathogen is also able to evade human immunity \nby antigenic variation [7]. In addition to that, P. falciparum exhibits extreme genetic diversity that \nallows it to adjust to various environments via population processes [8].\nImportance of Vaccine Development\nSignificant progress has been made in creating malaria vaccines, most notably the RTS,S/AS01 \nvaccine [9] that has been reported to decrease cases of clinical malaria (by approximately 56 to 60 \nper cent) and severe malaria (by approximately 47 to 60 per cent) in children aged 5-17 years one \nyear after vaccination [10]. Moreover, by providing herd immunity, the RTS,S//AS01 vaccine can \nalso diminish malaria transmission within populations [11]. Mathematical modeling estimates that \nwere the vaccine’s coverage to reach the levels attained during routine vaccination of children, \nimmense reduction in malaria deaths would be witnessed [12]. Such efficacy points to the \npossibility of halting malaria through immunisation as a complementary measure to transmission \ncontrol and treatment [13].\nDespite the encouraging trend, several outstanding issues remain to be resolved in the search for a \nuniversally efficacious malaria vaccine among them the cost of development [13]. Furthermore, \npublic awareness of, and acceptance of the vaccine is also critical in the ultimate successful roll-\nout. Communities within malaria-endemic districts have been reported to have a positive view of \nimmunization, with most caregivers responding that they will immunize children against malaria \n[14,15]. But misinformation and poor awareness can deter acceptance, making transparent \nmeasures critical to teach people on the benefits of the vaccine [15,16].\nMoreover, the RTS,S vaccine also has its effect on herd protection, where it can contribute to herd \nimmunity and reduce malaria transmission within populations [11]. Mathematical modeling \nestimates that where coverage of the vaccine is reached up to levels that have been achieved by \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n3\nroutine vaccinations of children, there can be significant reductions of malaria deaths and malaria \ncases [12]. Such possibilities reinforce the means by which malaria immunization can be a key \ncomponent of overall population interventions for malaria prevention and, ultimately, malaria \nelimination.\nRole of T-Cell Receptors in Immune Response\nT-cell receptors (TCRs) are of critical significance in malaria immunity, even more on P. \nfalciparum infection. The role of TCR is to recognize the antigens in the major histocompatibility \ncomplex, and hence induce cell activation thus resulting in immune responses [17]. The activation \nof CD4+ and CD8+ cell subsets has a critical role in malaria defense, where cells coordinate \ncellular and humoral immune responses capable of containing and clearing the parasite [18].\nThe CD4+ T cells, also referred to as helper T cells, are also key in the coordination of the immune \nresponse. These cells engage in activating B cells, generating antibodies, and activating the \ncytotoxic action of the CD8+ T cells, directly killing infested cells [19]. During malaria infection, \nthe CD4+ T cells can generate multi-lineages of cells, such as the Th1 cells, of major importance \nfor the control of malaria infection with Plasmodium via generation of pro-inflammatory cytokines \nsuch as IFN-γ and TNF-α [20]. The action of the Th1 cells has been associated with parasitemia \ncontrol and malaria infection outcomes. There also exists the Tr1 cells, a key subset of \nimmunosuppressive CD4+ T cells, whose action inhibits protection against the parasite by the \naction of the Th1 cells, and enhances generation of infection and inhibits immunological disease \nof malaria infection [21].\nMoreover, the dynamics of T-cells' response to malaria are also influenced by infection chronicity. \nRepeated antigenic malaria stimuli can lead to exhaustion of B-cells and also of T-cells. It is \ndemonstrated that frequent P. falciparum parasites exposure is followed by elevated levels of CD4 \nT-cells that exhibit phenotypic markers of exhaustiveness. It is evident on programmed cell death-\n1 (PD-1) alone and also co-expression of PD-1 along with lymphocyte-activation gene-3 (LAG-\n3). The proliferation of PD-1 and co-expression of PD-1/LAG-3 is of specific interest to CD45RA+ \nCD4 T-cells [22]. It is evident on animal models and also on humans where exhaustively \ndifferentiated T cells exhibit reduced cytokine generation along with proliferative activity [22]. It \nis important to understand these processes to determine means of enhancing responses of T-cells, \ne.g., by employing therapy that inhibits inhibitory pathways to restore function of T-cells.\nIn addition to CD4+ cells, CD8+ cells also make important contributions to malaria immunity. The \ncytotoxic cells can identify infected hepatocytes and red blood cells and destroy them, thereby \ninhibiting proliferation of the parasite [23,24]. It has been shown that malaria can be recognized \nby CD8+ cells through cross-presentation, enhancing their ability to respond to infection of the \nblood stage [25]. The activity of the CD8+ cells also relies on memory characteristics, and these \ncan be changed with history of infection, and with access of specific antigen [26].\nMolecular Docking and Immunoinformatics\nMolecular docking and immunoinformatics have been critical drug discovery tools. The \ntechnologies have contributed immensely in understanding how medicines combat some of the \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n4\nepidemic diseases such as COVID-19 and malaria. Molecular docking is a computer program \nwhich first made its entry in the 1980s in an effort of making it possible for addressing best pose \nof a target molecule when complexing a receptor [27]. The first program for molecular docking, \nhowever, was developed by Kuntz and his group in 1980s [28]. Molecular docking revolutionized \nunderstanding of functions of a molecule on a molecular scale [27]. It made it possible for \ncomplexation of a molecule and a receptor, calculation of affinities of complexation, and actual \nmedicine designing [29,30]. It accelerates drug discovery making it possible for screening of \nthousands of compounds for identifying a potential medicine and reducing effort and time in \nconfirming results in experiments [31].\nOn the other hand, immunoinformatics, or computational immunology, employs computational \nstrategies to analyze and predict immune reactions with a major focus on the design of vaccines. \nImmunoinformatics has recently emerged as an important tool in immunological studies, which \ninclude vaccine design and development. Despite that this technology is still evolving, the \ncomputational models have played a significant role during the selection of antigens or proteins \nand complex immunologic data analysis, thus facilitating the formulation of new testable \nhypotheses [32]. Immunoinformatics is established on the knowledge of the antigens' epitopes that \nare the targets of the immune response to develop multi-epitopic vaccines that can elicit strong \nimmune reactions [33]. One of the key areas of immunoinformatic applications is the prediction \nof B-and T-cell epitopes by computational strategies to aid the development of vaccines that can \nactivate adaptive immunity [34]. It is particularly relevant to emerging pathogens like the SARS-\nCoV-2 where classical strategies to the development of vaccines can fail [35]. The combination of \nimmunoinformatic analysis with molecular docking enables the rational peptide vaccine designing \nby predicting the affinities of the peptide-MHC interaction, their recognition by the major \nhistocompatibility complex (MHC) molecules, and by the T cells. These techniques allow for \nselection of best conformations that can be validated through experimental studies. The molecular \ndocking also enables screening of the potential vaccines to interact with the Toll-like receptors \n(TLRs), a major component of initiating the innate immunity [35]. Numerous studies have reported \nthat reformulating the key proteins used in the vaccine improves efficacy and immunogenicity thus \nimproving response and protection.\nComputational biology has gained popularity in the recent past and it has accelerated drug \ndiscovery and development. The latest technologies involving artificial intelligence (AI) and deep \nlearning have been widely used on research platforms, thus accelerating drug discovery. Deep \nlearning is based on the idea of artificial neural networks (ANNs) that can resemble the brain \nlearning process [36]. In most occasions, AI, that assists in predicting drug-targets, has \nsignificantly enhanced drug discovery by molecular docking and immunology. In this way, over a \nsingle conformation is obtained and the best of them are chosen depending on their binding \ncapacity. The strategy has helped in developing vaccines and drugs significantly.\nObjective of the study\nThis research was conducted with the purpose of applying molecular docking and \nimmunoinformatics to identify the most appropriate drug-target orientations. The study explicitly \ntakes into account the interaction of the antigen P. falciparum with the T-cell receptor in order to \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n5\nidentify the majority of the most appropriate malaria vaccine targets. Following the molecular \ndocking, binding affinities and interfaces of the complexes were determined in an effort to \ndetermine the activation of immune responses. Antigen-receptor docks were ranked and the most \nactive complexes, according to binding affinities, were chosen to undergo further investigations. \nThe results obtained in this study give a starting point of experimental testing and vaccine \ndevelopment in malaria.\nMethodology \nSelection of Antigens\nCriteria for Choosing Plasmodium falciparum Antigens\nThe P. falciparum antigens were chosen based on a range of factors related to their role in the \nlifecycle of the parasite and their status as vaccine targets. To achieve maximum coverage against \nmalaria, for example, priority was given to antigens with widespread sequence conservation \nbetween different P. falciparum strains. Also, our study prioritized antigens with key roles in \nimportant biological processes, such as host cell invasion, host cell egression, and immune evasion. \nThe reason for this prioritization was to maximize the probability of effective immune responses. \nAlso, our study focused on secreted and surface-expressed antigens. Surface-expressed and \nsecreted antigens are more likely targets for host immune systems. Based on previous studies, this \nstudy also considered the immunogenic potential of the antigens. Antigens that have shown \npromise in previous immunological studies (Table 1), eliciting strong immune responses in \nhumans or model organisms, were considered. \nTable 1: Summary of previous experimental studies validating Plasmodium falciparum antigens \nas vaccine candidates\nAntigen Key experimental \nstudy \nExperimental approach Main outcome\nPfCSP Mahmoudi et al., \n2017 [37]\nPhase III clinical trial in \nAfrican children\nDemonstrated relatively little \nefficacy towards malaria \ncontrol\nPfAARP Wickramarachchi et \nal., 2008 [38]\nImmunological \ncharacterization, sera from \nendemic area & invasion-\ninhibition assays\nPfAARP-N reactive with \nimmune sera and rabbit anti-\nAARP inhibited merozoite \ninvasion.\nPfRh5 Reddy et al., 2014 \n[39]\nRecombinant RH5 \nproduction and erythrocyte \nbinding assays\nRH5 binds erythrocytes \nsimilar to native protein - \nsupports vaccine candidacy.\nPfRipr Takashima et al., \n2022 [40]\nPfRipr fragment vaccine, \nimmunogenicity, and \ninhibitory antibodies\nPfRipr5 fragment induced \npotent invasion‐inhibitory \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n6\nantibodies against P. \nfalciparum blood stage.\nPfCyRPA Williams et al., \n2024 [41]\nBlood‐stage vaccine \ncomponent study with \nCyRPA, RH5, RIPR\nDemonstrated that CyRPA \n(with RH5/RIPR) is a \nconserved target and elicits \ninhibitory antibodies.\nPfMSP10 Bendezu et al., \n2019 [42]\nSerological assays in \nmalaria-exposed \nindividuals\nMSP10 found to be \nimmunoreactive; supports its \npotential as a vaccine or \nserological marker.\nPfSEA-1 Raj et al., 2014 [43] Animal immunization and \nin vitro egress assays\nAntibodies to PfSEA-1 reduce \nparasite replication by \narresting schizont rupture.\nSources of Antigen Sequences\nThe 3D structures of P. falciparum antigens were obtained from the Protein Data Bank (PDB) \n(https://www.rcsb.org/). The PDB is a database with structures of proteins that are experimentally \ndetermined using different techniques such as through X-ray crystallography or cryo-electron \nmicroscopy. However, some of the protein structures are not available the RCSB Protein Data \nBank and therefore their computational models were predicted using AlphaFold Structure \nDatabase (https://alphafold.ebi.ac.uk/). There are numerous antigen candidates that can be \nexplored for malaria vaccine development. Table 1 consists of some of the potential candidate \ntarget proteins that can be modelled for vaccine development. \nTable 2: Some of the potential Plasmodium falciparum antigens considered for vaccine development\nAntigen Source Identifier\nPfCSP (P. falciparum circumsporozoite protein) PDB 3VDJ\nPfSEA-1 (Schizont Egress Antigen 1) AlphaFold AF-A0A143ZXM2-F1\nPfAARP (Apical Asparagine-Rich Protein) AlphaFold AF-Q8IFN2-F1\nPfRh5 (Reticulocyte Binding Protein Homolog 5) PDB 4U0Q\nPfRipr (Rh5 Interacting Protein) PDB 8CDD\nPfMSP10 (Merozoite Surface Protein 10) AlphaFold AF-J7GNL4-F1\nPfCyRPA (Cysteine-Rich Protective Antigen) PDB 5EZN\nT-cell Receptor and MHC Molecules Data\nCriteria for Selection T-cell Receptor and MHC molecules\nIn this study, only TCRs known to interact with malaria-specific antigens or closely related \nepitopes were selected to ensure relevance to the P. falciparum  infection. To simulate human \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n7\nimmune responses, only TCR sequences from humans were chosen, as these receptors are crucial \nfor understanding potential vaccine efficacy and T-cell mediated immunity in humans. In addition, \npreference was given to TCRs with resolved crystal structures or high-quality modeled structures \nto enable accurate docking simulations. \nWhen choosing between TCR1 and TCR2 for docking simulation, it is important to consider their roles and \nprevalence. TCR1 (Gamma-Delta TCR) is less common type, found in about 1-5% of T cells, and \nis primarily located in mucosal tissues and skin. It has a function in localized immune reactions \nand is still not fully understood. TCR2 (Alpha-Beta TCR) is found in over 90% of T cells and \nparticipates in the majority of immune responses. It is the conventional TCR that interacts with \npeptide-MHC complexes to activate T cells [44]. For most studies, TCR2 (Alpha-Beta TCR) is the \npreferred choice due to its prevalence and well-characterized role in immune responses. \nMHC Class II molecules were selected in this study. For malaria vaccine design, MHC Class II \nmolecules are often studied. They stimulate helper T-cell and antibody-mediated immunity thus \nthey play a crucial role in generating a strong and sustained immune response, which is vital for \nlong-term protection against the parasite.\nDatabase and Sources for T-cell Receptor Structure and MHC molecules\nThe T-cell receptor (TCR) sequences and structures were retrieved from the Protein Data Bank \n(PDB) (https://www.rcsb.org/). The PDB was specifically used to obtain experimentally \ndetermined structures of human TCRs bound to various antigen-MHC complexes, particularly \nthose involved in recognizing malaria epitopes. In this study, TCR2 (Alpha-Beta TCR) (PDB ID: \n4WW1) was selected. MHC Class II was source from PDB (PDB ID: 3L6F)\nPreparation of Antigen and T-cell Receptor Structures\nProper preparation of protein structures is critical to the success of molecular docking studies. The \n3D structures of selected P. falciparum antigens and TCR structures were downloaded in PDB \nformat from the Protein Data Bank (PDB) and AlphaFold Structure Database. All water molecules, \nHetatm, and bound ligands were deleted from the protein structures using BIOVIA Discovery \nStudio Visualizer v24.1.0.23298.\nMolecular Docking Protocol\nThe Plasmodium antigens were docked to MHC Class II and further to TCR2 by means of ClusPro \nv2.0 server available on https://cluspro.org. It is a popular open source server for predicting diverse \nprotein-protein interaction. It was selected because it provides a stable protein-protein docking \nserver with a multi-step approach employing the use of the Fast Fourier Transform (FFT). Apart \nfrom that, the tool is capable of clustering the result based on binding energy and thus enable the \ndetermination of probable structures of the two protein bindings [45]. ClusPro is a web server \nwhose basic home page is used for general purposes. Only two files are accepted by the server in \nthe PDB format [46]. After the natural process of immune response, stepwise docking protocol \nwas conducted in this research in two steps. The antigens of P. falciparum, in step one, were \ndocked against Human MHC Class II protein. Once the initial docking was done, the binding \nenergies and compatibility were determined using the docking energy scores. The server could \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n8\nform complexes of various affinities and cluster populace. The docking pattern is a notable \nprocedure that is used in the prediction of affinities of the antigens to the MHC molecules \npreceding the presentation of the antigens to the T cells. This was formed with the assistance of \nnatural immune mechanisms whereby it conjugates the antigens with MHC molecules which are \nthen collected by the T cell receptors [47]. The most preferred MHC-antigen complexes that scored \nhighly were selected following successful phase one docking and docking to TCR2. \nIn a bid to test the possible recognition and activation of T cells, the pre-programming of antigen-\nMHC complex with TCR2 was done. The significance of MHC molecules is extremely high during \nthe presentation of processed plasmodium antigenic peptides to TCR2. This in turn triggers certain \nimmune reactions. The protocol permitted examination of each antigen-MHC complex's \npossibility for immunogenicity, simulating natural antigen recognition and presentation to T cells. \nDocking simulations for each were monitored through the ClusPro dashboard for any possible \ncomputational flaw. \nData Analysis\nDocking scores for all the complexes from ClusPro v2.0 were compared. Four energy parameter \nsets, namely, balanced, electrostatics favored, hydrophobic favored, and van der Waals + \nelectrostatics, were used to execute the models after successful docking. The models were then \nranked based on the cluster populations of the models, where a cluster's population represented the \nprobability of a given conformation. The model which belonged to the center of the biggest cluster \nof the hydrophobic-favored energy set was selected for analysis. The reason behind the selection \nof this model is that it has a large number of clusters and is biologically relevant as it is likely to \ngive an accurate depiction of the protein-protein interaction. The structural and functional integrity \nof the predicted complex was checked by visualizing and validating the chosen model with the \nhelp of BIOVIA Discovery Studio Visualizer. The models that were ranked highest were further \ninvestigated to determine the essential interactions that occurred at the antigen-TCR interface \nincluding hydrogen bonds, salt bridges and van der Waals forces. \nResults\nThe docking of proteins-proteins was done through ClusPro web platform and 30 models were \nproduced in each protein-protein complex. These models were classified into clusters according to \nstructural similarity and the clustering data were examined to give the most reasonable binding \nconformations. \nTable 3 presents the results of the clusters, cluster population, weighted score, and the minimum \nset of energy parameter. The Hydrophobic-favored parameter set of energy was prevalent in most \nof the models of interest, with the exception of the MHC-PfRh5-TCR apparently with the lowest \nenergy model with the van der Waals + Electrostatics (VdW+Elec) parameter set. Cluster \npopulations were diverse with the largest cluster size of MHC-PfCyRPA-TCR (174 members), \nMHC-PfMSP10-TCR (156 members) and MHC-PfCSP-TCR (154 members). It is interesting to \nnote that MHC-PfSEA1-TCR possessed the lowest weighted energy score (-1185.3) indicating a \nvery favorable binding conformation. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n9\nThe model of choice took into consideration the cluster size more than the energy scores because \nbigger clusters have a better chance of reflecting stable binding configurations [46]. Cluster 0 from \nthe Hydrophobic-favored Set was chosen for most complexes due to its high cluster population \nand relatively lower energy values. This aligns with previous observations that larger clusters \ncorrespond to energetically favorable binding landscapes. The PIPER energy Function for \nhydrophobic-favored and VdW+Elec are E = 0.40Erep + − 0.40Eatt + 600Eelec + 2.00EDARS and E \n= 0.40Erep + − 0.10Eatt + 600Eelec + 0.00EDARS respectively.\nTable 3: Summary of ClusPro Docking Results forAntigen-MHC-TCR Complexes\nComplex Cluster population Weighted Score Energy Parameter Set \n(Lowest Energy)\nMHC-PfSEA-1-TCR 104 -1185.3 Hydrophobic-favored\nMHC-PfRipr-TCR 112 -834.7 Hydrophobic-favored\nMHC-PfRh5-TCR 103 -265.3 VdW+Elec\nMHC-PfMSP10-TCR 156 -806.7 Hydrophobic-favored\nMHC-PfCyRPA-TCR 174 -948.4 Hydrophobic-favored\nMHC-PfCSP-TCR 154 -806.5 Hydrophobic-favored\nMHC-PfAARP-TCR 79 -1037.9 Hydrophobic-favored\nThe selected protein-protein complexes were further analyzed by visualizing their three-\ndimensional structures using BIOVIA Discovery Studio Visualizer. Figure 1 below shows P. \nfalciparum antigens in complex with MHC Class II and TCR2.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n10\n(a) (b)\n(c)\n(d) (e) (f) \n(g)\nFigure 1: Structural representations of the best-docked conformations of Pf antigens-MHC-TCR complexes visualized using BIOVIA Discovery Studio. The models correspond to \nthe cluster with high population and lowest-energy docked structures for Plasmodium antigens (a) PfAARP, (b) PfCSP, (c) CxRPA, (d) PfMSP10, (e) PfRh5, (f) PfSEA-1, and (g) \nPfRipr. The ribbon structures are color-coded to highlight secondary structural elements and interaction interfaces.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n11\nStructural analysis of the selected complexes was done using PDBsum \n(https://www.ebi.ac.uk/thornton-srv/databases/pdbsum/). The best 3 ranked antigen complexes \n(PfCyRPA, PfMSP10 and PfCSP) were analyzed for key interactions, including hydrophobic \ncontacts and hydrogen bonding at the interface. This helps in understanding the stability of the \ninteraction. \nTo assess the stereochemical quality and structural stability of the predicted complexes, \nRamachandran plot analysis was performed (Figure 2). According to standard validation criteria, \na high-quality model must have over 90% of residues in the most favored regions, on the basis of \nanalysis of 118 structures with a resolution of at least 2.0 Å and an R-factor of not more than 20.0.\n• PfCSP Complex: 82.7% of residues fall within the most favored regions, with 16.4% in \nadditionally allowed regions and only 0.5% in disallowed regions.\n• PfMSP10 Complex: 78.9% of residues are in the most favored regions, 19.5% in \nadditionally allowed regions, and 0.6% in disallowed regions.\n• CyRPA Complex: 77.3% of residues are within the most favored regions, 21.4% in \nadditionally allowed regions, and 0.3% in disallowed regions.\nWhile none of the models exceed the 90% threshold for the most favored regions, the presence of \na high percentage in additionally allowed regions suggests that the structures remain well-refined \nand acceptable for further analysis. The low percentage of disallowed residues indicates minimal \nsteric clashes, supporting the reliability of the predicted complexes.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n12\nPfCyRPA Complex\nPfMSP10 Complex\nPfCSP Complex \nFigure 2: An illustration of how different protein chains interact (column 1). Chains interact with each other using different type of interaction as indicated \nby colored lines. Each circle's size is proportional to the surface area of its corresponding protein chain. The area of the interface on each chain is shown \nas a colored wedge whose color is the same as the color of the other chain and whose size is proportional to the interface surface area. The Ramachandran \nplot (column 2) and statistics (column 3) for each complex shows the structural validation and overall stereochemical quality of the modeled \nstructures following protein-protein docking. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n13\nStatistics for all the interfaces are given below (Table 2, 3 and 4). The interaction analysis of the \ntop docked complexes reveals key stabilizing forces. In the PfCyRPA complex, non-bonded \ncontacts dominate, with a striking 467 interactions in the primary interface. Similarly, PfCSP \nexhibits 439 non-bonded contacts, while PfMSP10 shows 464. Hydrogen bonds are also notable, \nwith CyRPA forming 52 in its major interface, followed by PfMSP10 (46) and PfCSP (45). Among \nelectrostatic interactions, PfCyRPA stands out with 13 salt bridges, the highest among the three \ncomplexes. These results highlight the importance of non-bonded contacts and hydrogen bonds in \nstabilizing the complexes, with salt bridges playing a crucial role in PfCyRPA.\nTable 4: Interface statistics for PfCyRPA complex\nTable 5: Interface statistics for PfCSP complex\nTable 6: Interface statistics for PfMSP10 complex\nDiscussion \nAlthough the current study involves the application of computational docking to the study of \nantigen-receptor interactions, it is worth pointing out that some of the antigens used in the study \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n14\nare already known to be effective vaccine candidates that have been experimentally proven in prior \nstudies. Table 1 is a summary of published articles which give immunological or functional support \nof the vaccine-relevance of each antigen. An example is PfCSP vaccines (RTS,S/AS01), which \nhave demonstrated partial protection in children [48], PfAARP and PfRipr fragments produce \ninvasion-inhibitory antibodies and the RH5 -CyRPA -RIPR complex (including PfCyRPA) is a \nconserved blood-stage target. The consistency of our computational predictions with this existing \nbody of prior experimental evidence enhances the biological plausibility of the candidate antigens \nand gives a more explicit rationale in the validation of our hypothesis in subsequent in vivo and in \nvitro experiments.\nA comparison of our docking results with published immunoinformatics studies was done to put \nour findings into perspective. As an example, immunoinformatics vaccine candidates docked to \nTLR4 can easily generate dozens of docked poses (clusters) with binding energies of the order of \n-1,000 to -1,500 kcal/mol. One study found that a malaria multi-epitope vaccine, which was \ndocked to human TLR4, generated 30 clusters, the biggest cluster (cluster 0) containing 34 \nmembers and lowest-energy score of -1262.3 [49]. In another study, 30 models were generated \nand most optimal one contained energy -1514.2 (cluster center -1413.7) [50]. These values fall \nwithin the range of our values indicating that our docking energies are consistent with the \nprevious studies. Kozakov et al. [46] also noted that larger clusters typically suggest more \ncredible models.\nThe findings of this study give valuable insights into the interaction of human immune response \nwith P. falciparum antigens, of utmost significance for vaccine research. The critical antigenic \ninteractions were identified from the molecular docking simulations. These interactions can be of \nuse for future malaria vaccine design. This study yielded a total of seven complexes with different \ncluster population. \nOne of the significant outputs of this research is the strong binding interaction of PfCSP with MHC \nmolecule and T cell receptor. CSP is an important antigen that has been widely studied and has \nbeen well-characterized. It is promising and a key component of the novel RTS,S/AS01 malaria \nvaccine (Mosquirix). CSP is implicated at the beginning of P. falciparum infection to support the \nparasite’s invasion of the liver cells [51]. The high number of cluster population or members and \nthe low energy scores obtained in this study enhances the biological value of CSP as a target in the \nvaccine design. Incorporation of the CSP into the RTS,S vaccine has been reported to offer \nprotection against 56% of the case of the clinical malaria among 5 to 17-month-old infants [9]. \nThe outcome of this research also warrants the continuation of the research to develop and further \nimprove the effectiveness of the CSP-based vaccines.\nThe other key antigen that was studied is P. falciparum merozoite Rh5 interacting protein (PfRipr). \nThis antigen mostly associates with Rh5 and CyRPA during the invasion of the erythrocyte. \nDocking results revealed promising interaction between MHC, PfRipr, and TCR. This signifies its \npotential to serve as a strong immunogen target. According to recent studies, PfRipr has been fully \ncharacterized and it is a nonredundant protein. This protein plays a critical role in red blood cell \ninvasion. PfRipr is more conserved among proteins. These features make PfRipr an important \ncandidate especially for the creation of blood-stage malaria vaccines [52].\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n15\nBecause PfRipr happens in a complex with Rh5 and CyRPA, the interactions need to be included \nwithin the molecular docking simulations to preserve the native binding conformation [53] and \npreserve the natural sequence of interaction. Exclusion of molecules can potentially change the \nstructural stability of the binding site. This may impact the general efficacy of the candidate \nvaccine. The prospect of PfRipr as a future blood-stage vaccine target has already been shown by \na number of studies including Ntege et al., [54], Williams [55], Correia et al. [56] and Takashima \net al. [40]. Another key finding was noted with PfSEA-1. PfSEA-1 docking was found to give a \nsmall cluster with very low energy. A possible interpretation is that binding pose in PfSEA-1 is \nquite favorable yet distinct. Large cluster size is generally an indicator of confidence in ClusPro \n(when there are many similar low-energy poses) and a small cluster (even if low-energy) can \nindicate a lower count of repeated solutions. Kozakov et al. [46] note that cluster size is normally \na more useful ranking measure compared to raw energy. In this way, a small cluster with low \nenergy (such as PfSEA-1) cannot be considered as a large cluster.\nPfMSP10 is a significant protein located in the merozoite surface and apical end. It is a structurally \nmade up of two epidermal growth factors (EGF) where its C-terminal is attached to the red blood \ncell membrane. The role of PfMSP10 is still not known, but it is believed to contribute to growth \nstimulation and protein – protein interaction in merozoites and gametocytes [42]. The comparison \nof MSP1 with MSP10 in this study also puts into perspective the aspect of variability of the \nantigens to be remembered while developing vaccines. Although MSP1 is a well-studied protein \nwith a well-characterized target of the immune response, the extensive variability of the protein \namong P. falciparum parasites is a limitation to the potential as a vaccine candidate [57]. \nNevertheless, MSP10 was revealed to have strong interaction with the TCR with minimal \nvariability compared to the other proteins. It has been shown to be a stable protein to work with in \nterms of protein-protein interactions. MSP10 is also a potential serological marker and a potential \nvaccine candidate due to the strong induction of the immune response [42]. The outcomes of this \nstudy corroborate the utilization of MSP10 in future malaria vaccine research.\nIt is interesting to note that hydrophobic interactions usually prevail in protein-protein interfaces. \nThe structural analysis indicates that hydrophobic contacts are commonly used as binding \ninterfaces. As an example, in a T-cell receptor-peptide-HLA complex, all the 26 contacts \nbetween a Fab and a neo-antigenic peptide were hydrophobic [58]. This is similar to what we \nhave observed whereby clusters rich on hydrophobic contacts performed well. This is attributed \nto the fact that hydrophobic residues in the interface reduce solvent exposure and hence a major \ncontribution to binding free energy [58].\nProtein to protein binding affinity has been known to be enhanced by hydrophobic interactions. \nFrom our findings, it has been established that almost all peptide contacts are hydrophobic in \nantibody-peptide complexes [58]. Poses in the burying of hydrophobic surfaces in docking are \nfrequently energetically favourable in that, desolvation of nonpolar surfaces gives a large entropic \ngain. Therefore, clusters that have high hydrophobic interfaces are likely to dominate the scoring. \nPractically, numerous docking algorithms (such as ClusPro) will give clusters with a high number \nof hydrophobic contacts as the most populated or lowest-energy. The most populated clusters are \nlikely to represent those poses that have the largest area of hydrophobic interfaces, which is in \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n16\nagreement with the literature. This gives a biological explanation as to why the clusters which \n“hydrophobic-favored” appeared in our docking.\nThe molecular docking also provided hints on the stability of the complex structures of antigen-\nMHC-TCR. The Ramachandran plot also established the stability of the complex structures \npredicted, with a very high percentage of residues being in the most favored and also allowed \nregions. There were, however, no more than 90% of the residues being in the most favored regions \nfor the individual models, but the low percentage of disallowed residues (range of 0.3-0.6%) \nsuggests that the complex structures predicted are very well-refined and ready for other methods \nof analysis. The percentage of this structural validation ensures accuracy of computational \npredictions for vaccine design. The study of interaction identified stabilizing of the complex \nstructures with major contributions from non-bonded contacts and hydrogen bonds, with the \ncontribution of salt bridges being very prominent in PfCyRPA and PfMSP10. The observation \nconforms with earlier studies suggesting that hydrophobic and electrostatic interactions are very \nprominent in binding of antigens with TCR [46].\nLimitations of Docking-based Epitope Prediction\nThe prediction of epitopes using docking has drawbacks. It should be mentioned that docking \nfull-length antigens to MHC (or TLR) ignores important biological processes. To begin with, \npeptide cleavage is not performed. Generation of MHC-I epitopes in vivo is through the cleavage \nof proteins by proteasomes and their transportation by TAP [59]. The correct length and termini \nof peptides only come in the MHC groove. Proteasome cleavage and TAP efficiency prediction \nmethods such as NetCTL contain such tools due to the reason that it isolates which peptides are \nactually presented [59]. Our simulation ignored proteasomal processing, which may generate \npeptides that do not form in cells. As a matter of fact, the peptides have to be cut by proteasome \nand then delivered by TAP followed by MHC binding.\nSecond, the question of full-length and peptide docking might emerge. T-cells do not recognize \nfull length proteins [60], but only short peptides. According to Tong et al. [60], T cells perceive \nantigens in form of short peptide fragments in combination with MHC[6]. The docking of \ncomplete proteins to MHC is an approximate calculation, since the actual geometry is reliant \nupon a processed peptide bound to a part of the MHC groove. It is interesting to mention that our \ndocking does not replicate exactly the peptide - MHC interaction.\nThird, no refinement of molecular dynamics (MD) was performed. Docking is stiff and results in \na static snapshot. These complexes are frequently checked and refined with the help of MD \nsimulations. To illustrate, in a study where a vaccine was docked to TLR4, researchers observed \nthat following 50-100ns MD, the RMSD of the complex had stabilized, indicating that the \ninterface is stable [61]. We did not refined docked complexes using MD; where studies have \ndemonstrated that MD can uncover fluctuation or stabilize the interface.\nFinally, there was no HLA coverage analysis of the population. The efficacy of a vaccine is based \non the corresponding common HLA alleles. Contemporary vaccine designs actively apply \npopulation coverage technologies (e.g. IEDB) to make sure HLA is well represented. Using the \nIEDB population coverage tool, it has been demonstrated that some specific epitopes may have \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n17\nover 75 percent coverage of some local populations [62]. In our study, HLA allele coverage which \nis significant context of immunogenicity was not analyzed. We cannot guess what proportion of \nthe population our epitopes may guard without the analysis of coverage. As a matter of fact, peptide \nvaccine research comes out clearly to state the coverage on a global scale.\nConclusion\nThis study highlights the usefulness of computational docking in identifying P. falciparum \nantigens of excellent immunogenic potential as vaccine candidates. Among the complexes studied, \nPfCSP, PfRipr, and PfMSP10 showed stable and favorable interactions with human MHC–TCR. \nThis is in agreement with their recognized or emerging profile as successful vaccine candidates. \nImportantly, the MHC-PfSEA-1-TCR complex also scored the lowest energy and it suggests a \nhighly favorable binding conformation. However, its comparatively smaller population of cluster \nshows that the interaction may be less structurally stable and therefore may be the reason it did not \nemerge as a significant candidate. Nevertheless, the findings opens the way for further studies on \nthis antigen and in combination with other antigens. In addition, the findings point to the necessity \nof including multiple antigens of complementary function at different parasite life-cycle stages to \ndeal with problems of variability and immune evasion. The results broadly support the rationale \nfor the development of next-generation multi-epitope malaria vaccines, in which antigen selection \nneeds to be guided by binding affinity as well as conformational stability. Despite the existing \nexperimental validation, further targeted testing is still needed for the predicted complexes, \nparticularly PfSEA-1 and PfMSP10, for immunogenicity and protective efficacy. Through the \ncombination of computational prediction and experimental validation, this approach can accelerate \nthe development of more effective and broadly protective malaria vaccines.\nConflict of Interest\nThe authors declare no conflicts of interest.\nFunding Information\nThis research received no financial support from any organization\nReference \n[1] M. Van Den Berg, B. Ogutu, N.K. Sewankambo, N. Biller-Andorno, M. Tanner, RTS,S \nmalaria vaccine pilot studies: addressing the human realities in large-scale clinical trials, \nTrials 20 (2019) 316. https://doi.org/10.1186/s13063-019-3391-7.\n[2] CDC, Malaria’s Impact Worldwide, Malaria (2024). \nhttps://www.cdc.gov/malaria/php/impact/index.html (accessed November 5, 2025).\n[3] A.A. Koffi, S. Camara, L.P. Ahoua Alou, W.A. Oumbouke, R.Z. Wolie, I.Z. Tia, E.D. \nSternberg, F.H.A. Yapo, F.M. Koffi, S.B. Assi, J. Cook, M.B. Thomas, R. N’Guessan, \nAnopheles vector distribution and malaria transmission dynamics in Gbêkê region, central \nCôte d’Ivoire, Malar J 22 (2023) 192. https://doi.org/10.1186/s12936-023-04623-1.\n[4] W.L. Mandala, V. Harawa, F. Dzinjalamala, D. Tembo, The role of different components of \nthe immune system against Plasmodium falciparum malaria: Possible contribution towards \nmalaria vaccine development, Molecular and Biochemical Parasitology 246 (2021) 111425. \nhttps://doi.org/10.1016/j.molbiopara.2021.111425.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n18\n[5] S. Trivedi, A. Chakravarty, Neurological Complications of Malaria, Curr Neurol Neurosci \nRep 22 (2022) 499–513. https://doi.org/10.1007/s11910-022-01214-6\n[6] K.K. Zambare, A.B. Thalkari, N.S. Tour, A review on pathophysiology of malaria: a \noverview of etiology, life cycle of malarial parasite, clinical signs, diagnosis and \ncomplications, Asian Journal of Research in Pharmaceutical Science 9 (2019) 226.\n[7] T.N. Wells, P.L. Alonso, W.E. Gutteridge, New medicines to improve control and \ncontribute to the eradication of malaria, Nature Reviews Drug Discovery 8 (2009) 879–891.\n[8] S.I. Oyedeji, P.U. Bassi, S.A. Oyedeji, O. Ojurongbe, H.O. Awobode, Genetic diversity and \ncomplexity of Plasmodium falciparum infections in the microenvironment among siblings \nof the same household in North-Central Nigeria, Malar J 19 (2020) 338. \nhttps://doi.org/10.1186/s12936-020-03415-1.\n[9] M.B. Laurens, RTS,S/AS01 vaccine (Mosquirix TM): an overview, Hum Vaccin Immunother \n16 (2019) 480–489. https://doi.org/10.1080/21645515.2019.1669415.\n[10] J.K. Nunes, V. Cárdenas, C. Loucq, N. Maire, T. Smith, C. Shaffer, K. Måseide, A. Brooks, \nModeling the public health impact of malaria vaccines for developers and policymakers, \nBMC Infectious Diseases 13 (2013) 295. https://doi.org/10.1186/1471-2334-13-295.\n[11] M.A. Penny, R. Verity, C.A. Bever, C. Sauboin, K. Galactionova, S. Flasche, M.T. White, \nE.A. Wenger, N. Van de Velde, P. Pemberton-Ross, Public health impact and cost-\neffectiveness of the RTS, S/AS01 malaria vaccine: a systematic comparison of predictions \nfrom four mathematical models, The Lancet 387 (2016) 367–375.\n[12] M.T. White, R. Verity, T.S. Churcher, A.C. Ghani, Vaccine approaches to malaria control \nand elimination: Insights from mathematical models, Vaccine 33 (2015) 7544–7550. \nhttps://doi.org/10.1016/j.vaccine.2015.09.099\n[13] H. Mumtaz, A. Nadeem, W. Bilal, F. Ansar, S. Saleem, Q.A. Khan, T. Tango, C. Farkouh, \nN.F. Belay, R. Verma, M. Farkouh, M. Saqib, Acceptance, availability, and feasibility of \nRTS, S/AS01 malaria vaccine: A review, Immunity, Inflammation and Disease 11 (2023) \ne899. https://doi.org/10.1002/iid3.899.\n[14] U.M. Chukwuocha, P.C. Okorie, G.N. Iwuoha, S.N. Ibe, I.N. Dozie, B.E. Nwoke, \nAwareness, perceptions and intent to comply with the prospective malaria vaccine in parts \nof South Eastern Nigeria, Malaria Journal 17 (2018) 187. https://doi.org/10.1186/s12936-\n018-2335-0.\n[15] B.N. Emmanuel, A.N. Ishaq, O.Z. Akunne, U.F. Saidu, Evaluating the knowledge, attitude, \nperception, and readiness of caregivers of under 5-year-old children to accept malaria \nvaccine in Nigeria, Clinical and Experimental Vaccine Research 13 (2024) 121.\n[16] A. Meñaca, H. Tagbor, R. Adjei, C. Bart-Plange, Y. Collymore, A. Ba-Nguz, K. Mertes, A. \nBingham, Factors Likely to Affect Community Acceptance of a Malaria Vaccine in Two \nDistricts of Ghana: A Qualitative Study, (2014). \nhttps://doi.org/10.1371/journal.pone.0109707.\n[17] N.L. La Gruta, S. Gras, S.R. Daley, P.G. Thomas, J. Rossjohn, Understanding the drivers of \nMHC restriction of T cell receptors, Nature Reviews Immunology 18 (2018) 467–478.\n[18] Q. Li, Z. Ruan, H. Zhang, N. Peng, S. Zhao, L. Qin, X. Chen, Characterization of peripheral \nblood T lymphocyte subsets in Chinese rhesus macaques with repeated or long-term \ninfection with Plasmodium cynomolgi, Parasitology Research 110 (2011) 961–969. \nhttps://doi.org/10.1007/s00436-011-2581-3.\n[19] D.A. Rao, T cells that help B cells in chronically inflamed tissues, Frontiers in Immunology \n9 (2018) 1924.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n19\n[20] S.M. Muxel, A.P.F. do Rosário, C.A. Zago, S.I. Castillo-Méndez, L.R. Sardinha, S.M. \nRodriguez-Málaga, N.O.S. Câmara, J.M. Álvarez, M.R.D. Lima, The Spleen CD4+ T Cell \nResponse to Blood-Stage Plasmodium chabaudi Malaria Develops in Two Phases \nCharacterized by Different Properties, (2011). \nhttps://doi.org/10.1371/journal.pone.0022434.\n[21] R. Kumar, S. Ng, C. Engwerda, The role of IL-10 in malaria: a double edged sword, \nFrontiers in Immunology 10 (2019) 229.\n[22] J. Illingworth, N.S. Butler, S. Roetynck, J. Mwacharo, S.K. Pierce, P. Bejon, P.D. \nCrompton, K. Marsh, F.M. Ndungu, Chronic exposure to Plasmodium falciparum is \nassociated with phenotypic evidence of B and T cell exhaustion, The Journal of \nImmunology 190 (2013) 1038–1047.\n[23] C. Junqueira, C.R. Barbosa, P.A. Costa, A. Teixeira-Carvalho, G. Castro, S. Sen Santara, \nR.P. Barbosa, F. Dotiwala, D.B. Pereira, L.R. Antonelli, Cytotoxic CD8+ T cells recognize \nand kill Plasmodium vivax–infected reticulocytes, Nature Medicine 24 (2018) 1330–1336.\n[24] W.R. Weiss, S. Mellouk, R.A. Houghten, M. Sedegah, S. Kumar, M.F. Good, J.A. \nBerzofsky, L.H. Miller, S.L. Hoffman, Cytotoxic T cells recognize a peptide from the \ncircumsporozoite protein on malaria-infected hepatocytes., The Journal of Experimental \nMedicine 171 (1990) 763–773.\n[25] K. Yui, Cross ‐presentation of malaria antigen by brain microvessels: why CD 8 + T cells are \ncritical for the pathogenesis of experimental cerebral malaria, EMBO Mol Med 5 (2013) \n967–969. https://doi.org/10.1002/emmm.201302849.\n[26] A.D. Akue, J.-Y. Lee, S.C. Jameson, Derivation and Maintenance of Virtual Memory CD8 \nT Cells, The Journal of Immunology 188 (2012) 2516–2523. \nhttps://doi.org/10.4049/jimmunol.1102213.\n[27] X.-Y. Meng, H.-X. Zhang, M. Mezei, M. Cui, Molecular docking: a powerful approach for \nstructure-based drug discovery, Current Computer-Aided Drug Design 7 (2011) 146–157.\n[28] I.D. Kuntz, J.M. Blaney, S.J. Oatley, R. Langridge, T.E. Ferrin, A geometric approach to \nmacromolecule-ligand interactions, Journal of Molecular Biology 161 (1982) 269–288.\n[29] S.R. Arsad, H. Maarof, W.A. Wan Ibrahim, H.Y. Aboul‐Enein, Theoretical and Molecular \nDocking Study of Ketoconazole on Heptakis(2,3,6‐tri‐ O ‐methyl)‐β‐cyclodextrin as Chiral \nSelector, Chirality 28 (2016) 209–214. https://doi.org/10.1002/chir.22554.\n[30] I. Asiamah, S.A. Obiri, W. Tamekloe, F.A. Armah, L.S. Borquaye, Applications of \nmolecular docking in natural products-based drug discovery, Scientific African 20 (2023) \ne01593.\n[31] B. Zhang, H. Li, K. Yu, Z. Jin, Molecular docking-based computational platform for high-\nthroughput virtual screening, CCF Trans. HPC 4 (2022) 63–74. \nhttps://doi.org/10.1007/s42514-021-00086-5.\n[32] J.C. Tong, E.C. Ren, Immunoinformatics: current trends and future directions, Drug \nDiscovery Today 14 (2009) 684–689.\n[33] S. Banerjee, K. Majumder, G.J. Gutierrez, D. Gupta, B. Mittal, Immuno-informatics \napproach for multi-epitope vaccine designing against SARS-CoV-2, bioRxiv (2020) \n2020.07.23.218529. https://doi.org/10.1101/2020.07.23.218529.\n[34] F.A. Olotu, M.E.S. Soliman, Immunoinformatics prediction of potential B-cell and T-cell \nepitopes as effective vaccine candidates for eliciting immunogenic responses against \nEpstein–Barr virus, Biomed J 44 (2021) 317–337. https://doi.org/10.1016/j.bj.2020.01.002.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n20\n[35] T. Kar, U. Narsaria, S. Basak, D. Deb, F. Castiglione, D.M. Mueller, A.P. Srivastava, A \ncandidate multi-epitope vaccine against SARS-CoV-2, Scientific Reports 10 (2020) 10895.\n[36] A.M. Zador, A critique of pure learning and what artificial neural networks can learn from \nanimal brains, Nature Communications 10 (2019) 3770.\n[37] S. Mahmoudi, H. Keshavarz, Efficacy of phase 3 trial of RTS, S/AS01 malaria vaccine: \nThe need for an alternative development plan, Human Vaccines & Immunotherapeutics 13 \n(2017) 2098–2101. https://doi.org/10.1080/21645515.2017.1295906.\n[38] T. Wickramarachchi, Y.S. Devi, A. Mohmmed, V.S. Chauhan, Identification and \ncharacterization of a novel Plasmodium falciparum merozoite apical protein involved in \nerythrocyte binding and invasion, PloS One 3 (2008) e1732.\n[39] K.S. Reddy, A.K. Pandey, H. Singh, T. Sahar, A. Emmanuel, C.E. Chitnis, V.S. Chauhan, \nD. Gaur, Bacterially Expressed Full-Length Recombinant Plasmodium falciparum RH5 \nProtein Binds Erythrocytes and Elicits Potent Strain-Transcending Parasite-Neutralizing \nAntibodies, Infect Immun 82 (2014) 152–164. https://doi.org/10.1128/IAI.00970-13.\n[40] E. Takashima, H. Nagaoka, R. Correia, P.M. Alves, A. Roldão, D. Christensen, J.A. \nGuderian, A. Fukushima, N.K. Viebig, H. Depraetere, A novel asexual blood-stage malaria \nvaccine candidate: PfRipr5 formulated with human-use adjuvants induces potent growth \ninhibitory antibodies, Frontiers in Immunology 13 (2022) 1002430.\n[41] B.G. Williams, L.D. King, D. Pulido, D. Quinkert, A.M. Lias, S.E. Silk, R.J. Ragotte, H. \nDavies, J.R. Barrett, K. McHugh, Development of an improved blood-stage malaria vaccine \ntargeting the essential RH5-CyRPA-RIPR invasion complex, Nature Communications 15 \n(2024) 4857.\n[42] J. Bendezu, E. Villasis, S. Morales Ruiz, K. Garro, B. Infante, R. Gutierrez-Loli, P. \nRodríguez, M. Fernández-Díaz, D. Gamboa, K. Torres, Evaluation of Plasmodium \nfalciparum MSP10 and its development as a serological tool for the Peruvian Amazon \nregion, Malar J 18 (2019) 327. https://doi.org/10.1186/s12936-019-2959-8.\n[43] D.K. Raj, C.P. Nixon, C.E. Nixon, J.D. Dvorin, C.G. DiPetrillo, S. Pond-Tor, H.-W. Wu, \nG. Jolly, L. Pischel, A. Lu, I.C. Michelow, L. Cheng, S. Conteh, E.A. McDonald, S. \nAbsalon, S.E. Holte, J.F. Friedman, M. Fried, P.E. Duffy, J.D. Kurtis, Antibodies to \nPfSEA-1 block parasite egress from RBCs and protect against malaria infection, Science \n344 (2014) 871–877. https://doi.org/10.1126/science.1254417.\n[44] J. Wo, F. Zhang, Z. Li, C. Sun, W. Zhang, G. Sun, The Role of Gamma-Delta T Cells in \nDiseases of the Central Nervous System, Front. Immunol. 11 (2020). \nhttps://doi.org/10.3389/fimmu.2020.580304.\n[45] K.A. Porter, B. Xia, D. Beglov, T. Bohnuud, N. Alam, O. Schueler-Furman, D. Kozakov, \nClusPro PeptiDock: efficient global docking of peptide recognition motifs using FFT, \nBioinformatics 33 (2017) 3299–3301. https://doi.org/10.1093/bioinformatics/btx216.\n[46] D. Kozakov, D.R. Hall, B. Xia, K.A. Porter, D. Padhorny, C. Yueh, D. Beglov, S. Vajda, \nThe ClusPro web server for protein–protein docking, Nature Protocols 12 (2017) 255–278.\n[47] A. Johnson, J. Lewis, B. ALBERTS, Molecular biology of the cell, (2002). \nhttp://www.kubalibri.cz/files/9780393884852.pdf (accessed November 5, 2025).\n[48] P. Vandoolaeghe, L. Schuerman, The RTS,S/AS01 malaria vaccine in children 5 to 17 \nmonths of age at first vaccination, Expert Review of Vaccines 15 (2016) 1481–1493. \nhttps://doi.org/10.1080/14760584.2016.1236689.\n[49] A.A. Larbi, C. Mensah, R. Korankye, A.A. Darkwah, E.B. Agyemang, I.Y. Ayikpah, A. \nKwarteng, Immunoinformatics‐Based Multi‐Epitope Vaccine Design Against P. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n21\nfalciparum‐Causing Malaria: A Computational Approach, Health Sci Rep 8 (2025) e71353. \nhttps://doi.org/10.1002/hsr2.71353.\n[50] P. Bhalerao, S. Singh, V.K. Prajapati, T.K. Bhatt, Exploring malaria parasite surface \nproteins to devise highly immunogenic multi-epitope subunit vaccine for Plasmodium \nfalciparum, J Genet Eng Biotechnol 22 (2024) 100377. \nhttps://doi.org/10.1016/j.jgeb.2024.100377.\n[51] J. Zhao, P. Bhanot, J. Hu, Q. Wang, A Comprehensive Analysis of Plasmodium \nCircumsporozoite Protein Binding to Hepatocytes, PLoS One 11 (2016) e0161607. \nhttps://doi.org/10.1371/journal.pone.0161607.\n[52] A.J. Laurenson, E. Stucke, R. Scalsky, M.B. Laurens, In Silico Prediction of Plasmodium \nfalciparum PfRipr Epitopes as Vaccine Candidates, Journal of the Pediatric Infectious \nDiseases Society 12 (2023) S5–S6.\n[53] J.C. Volz, A. Yap, X. Sisquella, J.K. Thompson, N.T. Lim, L.W. Whitehead, L. Chen, M. \nLampe, W.-H. Tham, D. Wilson, Essential role of the PfRh5/PfRipr/CyRPA complex \nduring Plasmodium falciparum invasion of erythrocytes, Cell Host & Microbe 20 (2016) \n60–71.\n[54] E.H. Ntege, N. Arisue, D. Ito, T. Hasegawa, N.M.Q. Palacpac, T.G. Egwang, T. Horii, E. \nTakashima, T. Tsuboi, Identification of Plasmodium falciparum reticulocyte binding \nprotein homologue 5-interacting protein, PfRipr, as a highly conserved blood-stage malaria \nvaccine candidate, Vaccine 34 (2016) 5612–5622. \nhttps://doi.org/10.1016/j.vaccine.2016.09.028.\n[55] B.G. Williams, Investigation of the Plasmodium falciparum RH5-Interacting Protein \n(PfRIPR) as a blood-stage malaria vaccine target, PhD Thesis, University of Oxford, 2021. \nhttps://ora.ox.ac.uk/objects/uuid:387bda94-1500-4dc8-94be-2e50febd5cc4 (accessed \nNovember 5, 2025).\n[56] R. Correia, B. Fernandes, R. Castro, H. Nagaoka, E. Takashima, T. Tsuboi, A. Fukushima, \nN.K. Viebig, H. Depraetere, P.M. Alves, Asexual blood-stage malaria vaccine candidate \nPfRipr5: Enhanced production in insect cells, Frontiers in Bioengineering and \nBiotechnology 10 (2022) 908509.\n[57] C. Putaporntip, N. Kuamsab, R. Rojrung, S. Seethamchai, S. Jongwutiwes, Structural \norganization and sequence diversity of the complete nucleotide sequence encoding the \nPlasmodium malariae merozoite surface protein-1, Scientific Reports 12 (2022) 15591.\n[58] K.M. Wright, S.R. DiNapoli, M.S. Miller, P. Aitana Azurmendi, X. Zhao, Z. Yu, M. \nChakrabarti, W. Shi, J. Douglass, M.S. Hwang, E.H.-C. Hsiue, B.J. Mog, A.H. Pearlman, S. \nPaul, M.F. Konig, D.M. Pardoll, C. Bettegowda, N. Papadopoulos, K.W. Kinzler, B. \nVogelstein, S. Zhou, S.B. Gabelli, Hydrophobic interactions dominate the recognition of a \nKRAS G12V neoantigen, Nat Commun 14 (2023) 5063. https://doi.org/10.1038/s41467-\n023-40821-w.\n[59] J. Sarvmeili, B. Baghban Kohnehrouz, A. Gholizadeh, D. Shanehbandi, H. Ofoghi, \nImmunoinformatics design of a structural proteins driven multi-epitope candidate vaccine \nagainst different SARS-CoV-2 variants based on fynomer, Sci Rep 14 (2024) 10297. \nhttps://doi.org/10.1038/s41598-024-61025-2.\n[60] J.C. Tong, E.C. Ren, Immunoinformatics: Current trends and future directions, Drug \nDiscov Today 14 (2009) 684–689. https://doi.org/10.1016/j.drudis.2009.04.001.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint \n\n22\n[61] S. Banerjee, K. Majumder, G.J. Gutierrez, D. Gupta, B. Mittal, Immuno-informatics \napproach for multi-epitope vaccine designing against SARS-CoV-2, bioRxiv (2020) \n2020.07.23.218529. https://doi.org/10.1101/2020.07.23.218529.\n[62] A.D. Sharma, J.S.L. Magdaleno, H. Singh, A.F.C. Orduz, L. Cavallo, M. Chawla, \nImmunoinformatics-driven design of a multi-epitope vaccine targeting neonatal rotavirus \nwith focus on outer capsid proteins VP4 and VP7 and non structural proteins NSP2 and \nNSP5, Sci Rep 15 (2025) 11879. https://doi.org/10.1038/s41598-025-95256-8.\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 20, 2026. ; https://doi.org/10.64898/2026.03.18.712575doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}