Computational Analysis of Antimicrobial Peptides Targeting Agglutinin-Like Protein 3 (Als3) and β-Glucan Synthase in Denture-Associated Candidiasis Authors: Ravinder S Saini, Doni Dermawan, Shubham Chopra, Rayan Ibrahim H Binduhayyim, Seyed Ali Mosaddad, Mario Alberto Alarcón-Sánchez, Artak Heboyan Published: 2026 ABSTRACT AbstractBackgroundCandida albicans adhesion and biofilm formation on prosthetic surfaces drive denture-associated candidiasis and are unaffected by traditional fungicidal treatments. Pre-emptive prevention of fungal attachment to prosthetic devices is a rational strategy to avoid recurrent infections.MethodsWe utilized a mechanism-based computational pipeline to screen peptides with activity centred on inhibiting fungal adhesion instead of viability. We chose two colonization-relevant targets: agglutinin-like protein 3 (Als3), involved in surface attachment, and 1,3-β-glucan synthase (GS), involved in glucan-dependent biofilm-support processes. Peptides were ranked using protein–peptide docking, 100-ns molecular dynamics simulations, and MM/PBSA binding free energy calculations to determine metrics of interface coverage, interaction persistence, and dynamic energetic favourability.ResultsWe found that tachystatin consistently and stably occupied the highest proportion of adhesion-related interfaces on both Als3 and GS. Tachystatin dynamics demonstrated lower residue-level fluctuations in functionally relevant residues and preservation of hydrogen bonds throughout the simulation. MM/PBSA assessments revealed favourable and consistent binding energetics facilitated by van der Waals forces. These findings suggest that the peptide may comparatively maintain interface occupancy consistent with a potential steric hindrance effect at adhesion-relevant regions. The selected control peptides failed to occupy the predicted interfaces and demonstrated inadequate stabilities.ConclusionsTargeting Als3-mediated adhesion together with GS-associated biofilm-support processes represent a promising preventive strategy for denture-associated candidiasis. Tachystatin emerged as the most promising peptide within this comparative computational screen and may warrant further evaluation as a candidate for reducing biofilm formation on dental prosthetics Overall, this interaction- and simulation-based framework may support antimicrobial peptide selection focused on biofilm prevention.Key words: Als3; 1,3-β-glucan synthase; Dentures; Candidiasis; Molecular docking Introduction Denture stomatitis or denture-associated candidiasis (DAC) is one of the most common problems encountered among removable dental prosthesis users worldwide. Elderly and medically compromised patients show high rates of colonization by Candida albicans. DAC can result in chronic inflammation of tissues supporting dentures, erythema of the mucosa covered by the denture and painful recurrent denture stomatitis.1 Clinical persistence and recurrence of fungal infections in prosthetic devices results from stable organism attachment and biofilm development instead of rampant fungal growth because biofilms protect fungal cells from both mechanical and chemical removal methods.2,3 Traditional antifungals, such as azoles and polyenes, work by disrupting fungal metabolism or fungal cell membranes. Traditional antifungal agents can prevent the growth of planktonic Candida cells; however, limited success has been observed when treating surface-attached biofilms due to limited drug penetration, metabolic changes, and stress tolerance mechanisms, resulting in biofilm persistence.4 Without consistent mechanical disruption of Candida biofilms on dental prostheses (an environment constantly being recolonized with Candida and other microbes), fungicidal therapy may not be sufficient to eliminate Candida biofilms. Therefore, preventing the adhesion of Candida may be more beneficial than attempting to kill Candida once it has been established. The binding of Candida to prosthetic surfaces is protein dependent. Molecular components responsible for adhesion include Agglutinin-like protein 3 (Als3), which is primarily involved in binding to host cells and non-living surfaces, and 1,3-β-glucan synthase (GS), which aids in biofilm anchoring and formation through its role in cell wall production.5, 6, 7 Since Als3 is involved in early adhesion and GS in downstream biofilm-support processes, these proteins provide complementary colonization-relevant targets for evaluating an adhesion-focused preventive strategy.8 Targeting the early stages of biofilm formation may inhibit the later stages, reducing the fungal burden without necessarily causing high levels of resistance. Antimicrobial peptides (AMPs) are natural components of innate immunity that have gained increased attention as potential replacements for traditional antifungals. In contrast to small molecules, AMPs are characterized by many overlapping mechanisms of action, including membrane perturbation, surface binding, and interference with protein–protein interactions.9 Mechanistically relevant to prosthetics, some AMPs display surface-active and adhesive modulation activity, which implies that AMPs benefits may not directly correlate with fungicidal efficacy.10 Nevertheless, most computational/experimental studies assess AMPs based on binding affinity or membrane lysis without considering if peptide–protein binding disrupts important adhesion dependent processes.11 One weakness of many AMPs identification campaigns is the lack of target-based selection strategies that can differentiate between peptides that can interfere with fungal attachment and those that only demonstrate temporary binding or cell toxicity.12 The results of extensive peptide screens targeting diverse fungal species do not typically provide meaningful information for translation or materials science applications, such as dental coatings or delivery systems at the point of use.13 To function as an AMPs for prosthetic-related infections, the peptide must secure stable binding positions and persist through fluid conditions while being suitable for distribution on surfaces. Docking affinity measurements alone do not fulfil the requirements for AMP effectiveness. Therefore, we implemented a targeted, mechanism-based computational approach to screen for AMPs that could potentially interfere with Candida adhesion to denture surfaces. Targeting Als3 together with GS allowed us to examine whether peptide binding could interfere with two colonization-relevant stages, that is initial adhesion and subsequent biofilm-support processes.14,15 Molecular docking and dynamics simulations, along with MM/PBSA binding free energy calculations, were used to determine the stability of interactions, residue-specific interactions, and retention of conformations. Instead of focusing on peptide diversity, we focused on the functional applicability, mechanistic understanding, and translatability of this study to provide a rational approach for selecting AMPs as candidates for denture-targeted antifungal therapy. The pathogenesis of denture-induced Candida infection and our rationale for selecting Als3 and GS as complementary colonization-relevant targets are shown in Figure 1. Adhesion-driven pathogenesis of denture-associated candidiasis and rationale for selecting Als3 and GS as complementary colonization-relevant targets. Attachment of fungi to prosthetic surfaces is first facilitated by Als3, whereas subsequent biofilm-support processes are reinforced by 1,3-β-glucan synthase (GS). The highlighted areas indicate the interfaces evaluated in this study. Molecular modelling suggests that tachystatin may remain associated with selected adhesion-critical interfaces during the simulated timeframe, consistent with a potential role in hindering initial attachment and later biofilm stabilization. Materials and methods Study design and methodological framework In this study, a computational pipeline guided by a mechanism was implemented to screen AMPs with the potential to target C. albicans adhesion rather than survival. The process considered biological feasibility, longevity of the interaction, and clinical translatability about a dental prosthetic surface. Structural filtration through docking served as the first step in selection, which then underwent MD simulations and binding free energy analyses to identify stable, biologically relevant peptides–protein interactions at adhesion-related binding sites.16 Selection and preparation of adhesion-relevant Candida protein targets Targets were chosen based on a mechanism-based approach aimed at fungal adherence and initial biofilm formation on prosthetic surfaces. Instead of testing panels of un-associated fungal proteins, this study focused on proteins that facilitate C. albicans adherence and biofilm tethering, two factors largely responsible for denture-related candidiasis. Two colonization-relevant Candida proteins were selected for detailed analysis. Als3 was selected as the primary target based on its role as a cell surface adhesin involved in the initial attachment of C. albicans cells to host and abiotic surfaces, including dental prosthetic materials. Als3 is expressed during the early stages of colonization and mediates robust surface attachment; therefore, it is a biologically relevant target for the modulation of initial fungal attachment. GS known for its role in cell wall biogenesis and structural support of the glucan-dependent biofilm matrix was chosen as a secondary colonization-relevant target rather than as a classical surface adhesin. GS is not an obvious primary adhesin like Als3. It supports glucan production and biofilm structural integrity. Thus, it was felt to have relevance to early surface-associated persistence following initial attachment. GS was therefore chosen to provide representation of a downstream biofilm-support process that could aid in stabilization of Candida colonization on prosthetic surfaces. Als3 and GS were used in this study as complementary colonization-relevant targets representing distinct stages of prosthetic surface establishment. Als3 for initial adhesion and GS for early biofilm-support and matrix-associated persistence. This pairing was intended to provide a mechanistically broader framework for peptide screening while recognizing that the two proteins do not play equivalent biological roles in fungal adhesion. High-resolution solved structures of Als3 and GS were acquired from the Protein Data Bank.17 Protein preparation was performed using Swiss-PDB Viewer, where water molecules, heteroatoms, and crystallographic waters were deleted; missing side chains were added if applicable, hydrogen bonding networks were optimized, and energy minimization was performed to eliminate steric strain.18 CASTp 3.0 was used to identify surface grooves and functionally relevant pockets in conjunction with literature-defined functional regions. These prioritized regions were then used to define interface-focused docking targets and subsequent molecular dynamics simulations.19 The selected Candida targets, their roles in prosthetic colonization, the rationale for their inclusion, and the infection stage represented by each target are summarized in Table 1. Table 1Colonization-relevant Candida protein targets selected for mechanistic analysis.Table 1 dummy alt textProteinBiological role in prosthetic colonizationRationale for selectionStage of infectionAls3Surface adhesin mediating initial attachmentDirectly governs adhesion to abiotic surfacesEarly1,3-β-glucan synthase (GS)Glucan-dependent biofilm matrix support and cell wall biogenesisRepresents a downstream biofilm-support process relevant to early surface-associated persistencePost-attachment / early biofilm support Colonization-relevant Candida protein targets selected for mechanistic analysis. Mechanism-driven selection of antimicrobial peptides Antimicrobial peptides were selected using a stepwise mechanism-based approach that involved an initial broader screening stage. This was followed by focused comparative evaluation of a smaller set of mechanistically distinct peptides. The criteria included structural availability, peptide size allowing for surface delivery, amphipathic nature, and previous evidence of antifungal or surface-active activity. Following the initial screening stage three peptides were retained for primary comparative analysis in the main workflow. Tachystatin was prioritized for detailed analysis because the initial screening indicated favourable interface occupancy and interaction stability against the selected colonization-relevant targets. Magainin was included as a mechanistically relevant comparator with known antifungal membrane-disruptive activity. This allowed comparison between a broadly antifungal peptide and a peptide prioritized for adhesion-interface engagement. Peptide 35409 was included as a non-selective reference comparator. It has reported antimicrobial activity. It was not specifically selected for adhesion-targeted or biofilm-associated binding behaviour. The final comparative set was designed to contrast target-prioritized binding, general antifungal activity and non-selective reference behaviour within a focused mechanistic framework. The broader peptide screening pool along with docking-based prioritization data for additional peptide–protein systems is provided in the supplementary material to clarify the basis for narrowing the final comparative set. Molecular docking simulations Protein–peptide docking calculations were conducted through the High Ambiguity Driven protein–protein DOCKing (HADDOCK) webserver. Instead of allowing docking across the entire protein surface, docking was biased towards predefined regions on Als3 and GS that are relevant for colonization20 Functional residues supported by literature were used to identify important regions on each protein together with CASTp-derived surface grooves and pockets. These were hypothesized to play a role in initial adherence or early-stage biofilm-support mechanisms. Active residues were chosen from these prioritized functional regions. Solvent-accessible neighbouring residues were selected as passive residues to help bias the docking towards the interface. Complexes obtained from docking were ranked by HADDOCK scores, which take into account van der Waals, electrostatic, and desolvation energies. Docking was only used to help filter structures for later comparison. It is not a stand-alone predictor of biological function. Only docks where the peptide could plausibly occupy the target interface were chosen for subsequent molecular dynamics-based comparison.21 For methodological transparency, we summarize target proteins, interface-centric residue definitions used, and downstream criteria used to select proteins for progression into MD analysis in both the main text and Supplementary Table. Molecular dynamics simulations MD simulations were performed to determine the persistence of interactions, conformational stability, and occupation of the interface under physiological conditions. MD simulations were performed in GROMACS (version 2022.5) using the OPLS-AA/L force field. It was selected because it is a widely used all-atom force field for biomolecular simulations and provides a consistent framework for comparative evaluation of peptide–protein interactions under the same simulation conditions.17 The complexes were placed in an octahedral box, solvated with the SPC/E water model, and neutralized with counterions. Energy minimization was performed using the steepest descent algorithm. After minimization, the system was equilibrated under the NVT and NPT ensembles at 300 K and 1 bar pressure. After equilibration, production MD simulations were performed for 100 ns with periodic boundary conditions. All complexes were simulated using the same protocol to enable direct comparative interpretation across the screened peptide–protein systems. The structural stability of the complexes and interactions was analysed using the root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration, hydrogen bonds, and total potential energy. The analyses focused on adhesion interface stability rather than global protein movement. A single 100-ns production trajectory was generated for each selected complex and used as a comparative screening-level assessment of interaction persistence under identical simulation conditions. We recognize that replicate simulations and alternative starting conformations would provide more robust conformational sampling and stronger statistical support. The present MD analysis should be interpreted as comparative and hypothesis-generating rather than definitive. Binding free energy calculations To measure the strength of binding of each antimicrobial peptide to the adhesion-important regions of Als3 and GS, we calculated the binding free energies. MM/PBSA was employed as a computationally inexpensive post-simulation approach . This was done to obtain a comparative estimate of peptide–protein binding favourability from the flexible, equilibrated portions of the MD trajectories . It was not intended to rely on static docking poses alone. MM/PBSA was used herein strictly as a relative ranking tool to screen for persistence of interactions across complexes simulated in a uniform fashion rather than trying to generate absolute binding affinities. Snapshots were captured from the last equilibrated 20 ns of each 100 ns molecular dynamics trajectory to avoid capturing interactions during brief conformational changes. Binding free energies were computed with gmx_MMPBSA and solvation was calculated with Poisson–Boltzmann. The total binding free energy was computed according to the following expression: Where G represents the molecular mechanics energy combined with polar and nonpolar solvation contributions for the peptide–protein complex, isolated protein, and isolated peptide, respectively. The MM/PBSA results were interpreted comparatively with emphasis on relative energetic trends across the tested systems rather than on absolute affinity values. This approach is appropriate for screening-level prioritization. It also recognizes the known approximations of endpoint free-energy methods. We focused on the contribution of the van der Waals force, as a longer duration of steric occupancy on the adhesion surfaces mechanistically blocks fungal adhesion and biofilm formation. We also sought to distinguish between peptides showing predominantly transient electrostatic association and those demonstrating relatively more persistent and energetically favourable interface occupancy within this comparative computational framework. Interaction analysis and visualization Residue-residue interactions, such as hydrogen bonding, hydrophobic interactions, and residue-level steric blocking of adhesion grooves, were also computed along the MD trajectories. Conserved interaction hotspots were determined based on the contact frequency and spatial conservation of the contacts. Structural figures were generated and interactions mapped using PyMOL and UCSF Chimera for mechanistic insight.22 Methodological outcome Using this workflow, we identified peptide candidates predicted to occupy selected Candida binding interfaces in a mechanism-informed manner . This helped prioritize molecules for future evaluation in implant-directed antifungal applications, including coatings or point-of-use strategies. Results Structural relevance of Als3 and GS as adhesion-critical targets Als3 and GS were assessed as complementary colonization-relevant targets for the inhibition of early attachment and biofilm-support processes on prosthetic devices. Structural considerations indicated that Als3 contains a solvent-exposed ligand-binding groove that is rich in polar and nonpolar residues critical for binding to surfaces. GS has large surface patches involved in cell wall synthesis and biofilm binding to surfaces. These surface patches are biologically significant, as prolonged binding to the surface hinders initial binding and subsequent biofilm formation. Further investigation was performed into peptide binding and retention to these surfaces instead of whole protein binding. Docking identifies tachystatin as the most effective adhesion-interface binder Docking simulations were used as an interface-focused structural filtering step to evaluate whether the chosen antimicrobial peptides could plausibly occupy the predefined colonization-relevant regions of Als3 and GS. Among the screened peptides, tachystatin exhibited the most energetically favourable binding poses, deeply embedding itself into conserved patches on Als3 that are known to be important for fungal adherence. Tachystatin presented the largest interface and buried surface area within the binding grooves formed by residues known to be involved in binding, hydrophobic pockets, and key polar interactions. Magainin had an intermediate sized interface, and binding poses appeared to be shallower and more limited. As a positive control, peptide 35409, which does not specifically bind these surfaces, was unable to burrow deeply within grooves associated with binding and made less contact with residues known to be functionally important. Examples of docking poses demonstrating peptide interactions with Als3 and GS adhesion surfaces are shown in Figure 2. The data suggest that tachystatin may bind to these regions through spatial complementarity thereby supporting its prioritization for future experimental validation. Docking-derived binding modes of antimicrobial peptides at selected colonization-relevant interfaces of Als3 and GS. Molecular dynamics simulations confirm persistent adhesion-interface occupation The persistence of the predicted interactions under dynamic, physiologically relevant conditions was assessed by performing 100-ns molecular dynamics simulations of complexes containing the top-scoring peptides. The Tachystatin–Als3 and Tachystatin–GS complexes remained stable over the course of the simulation trajectories, with the RMSD remaining low and consistent between the targets (Figure 3A). Additionally, Tachystatin binding appears to restrict residue-level flexibility within the adhesion-important loops of both Als3 and GS, as determined by RMSF analysis (Figure 3B), suggesting that Tachystatin locally stabilizes functional regions. Conversely, fluctuations at the interface increased upon magainin binding. We also found that the Tachystatin–protein interaction did not cause an expanded protein conformation upon binding, as measured by the radius of gyration. As depicted over time in Figure 3 these findings suggest that tachystatin maintains comparatively stable occupation of the evaluated interfaces over the simulated timeframe. This is consistent with a potential adhesion-blocking effect. Molecular dynamics stability of peptide–protein complexes at adhesion interfaces. Hydrogen bonding and contact analysis reveal mechanistic adhesion blockade Interaction analysis revealed insights into the mechanism of adhesion inhibition. Tachystatin engaged in extensive and stable hydrogen bonding interactions with the adhesion residues of Als3 and GS, which persisted for significant amounts of time along the trajectories. Tachystatin also participates in many charged–apolar and apolar–apolar interactions, facilitating strong van der Waals contacts at the interface. This was accompanied by magainin forming fewer stable hydrogen bonds and more transient polar contacts than the control peptide, which had fewer interactions with residues important for adhesion (Figure 3). Additionally, when we examined the surface area coverage of adhesion sites through NIS analysis, we found that tachystatin contacted a wider surface area (Figure 4), which aligns with displacement as the occlusion of binding sites necessary for fungal adhesion. Interaction networks depicting these properties are shown in Figure 5. The data suggest that the predicted inhibitory effect may be more closely associated with prolonged interface occupation than with electrostatic attraction alone. Quantitative surface coverage of selected colonization-relevant regions on Als3 and 1,3-β-glucan synthase (GS) by antimicrobial peptides. Interaction networks supporting the predicted interface occupation of tachystatin at selected colonization-relevant targets. Energetic persistence favours tachystatin over comparator peptides MM/PBSA calculations of binding free energies were conducted to assess the energetic contribution of peptide occupancy at the adhesion determining interface. Tachystatin demonstrated the lowest (highest negative value) and least variable binding free energies when targeting Als3 and GS (values ranging from −58.2 to −93.5 kcal/mol) compared to magainin and the reference peptide. Energy decomposition analyses showed a greater van der Waals contribution for tachystatin. This suggests that non-covalent contacts may contribute importantly to its comparatively favourable interface persistence within this computational framework. Magainin and the reference peptide had lower persistence which was reflected by weaker stability during the MD runs. Table 2 summarizes the MD persistence data along with the MM/PBSA binding free energies for the representative peptide–protein complexes. Table 2 shows the correlation between persistence and energetic stability. Table 2Summary of dynamic stability and energetic persistence of peptide–protein complexes.Table 2 dummy alt textComplexAvg RMSD (Å)Avg H-bondsΔG binding (kcal/mol)Stability interpretationAls3–TachystatinRMSD 2.882H-bonds 27ΔG −76.59Stable adhesion blockadeGS–TachystatinRMSD 2.206H-bonds 36ΔG −42.40Persistent interface occupationAls3–MagaininRMSD 2.932H-bonds 25ΔG −42.40Partial adhesion interferenceControl peptideRMSD 2.815H-bonds 24ΔG −52.46Transient binding Summary of dynamic stability and energetic persistence of peptide–protein complexes. Integrated ranking of peptides based on adhesion-blocking potential We ranked the peptides according to interface occupation, dynamic stability, interaction persistence and binding free energy. Within this comparative computational framework tachystatin showed the most favourable overall profile among the evaluated peptides. Magainin demonstrated intermediate behaviour with lower interface persistence. The reference comparator showed weaker overall support for sustained interface engagement. Tachystatin was prioritized as the most promising candidate from this screened set for future investigation of adhesion-blocking effects on denture-relevant Candida colonization. Table 3 summarizes the antimicrobial peptides ranked by adhesion interface occupation, dynamic stability, persistence and predicted binding free energy. Table 3 summarizes all the antimicrobial peptides ranked by the percentage occupation of the adhesion interface, percentage dynamic stability, persistence, and predicted binding free energy. Table 3Integrated comparative ranking of antimicrobial peptides based on interface occupation, MD stability, and MM/PBSA-derived persistence.Table 3 dummy alt textPeptideAdhesion interface occupation (Als3 + GS)MD stabilityBinding persistence (MM/PBSA)Overall rankTachystatinMost favourableMost favourableMost favourable1 (most favourable in current comparative screen)MagaininIntermediateIntermediateIntermediate2Peptide 35409Less favourableLess favourableLess favourable3 Integrated comparative ranking of antimicrobial peptides based on interface occupation, MD stability, and MM/PBSA-derived persistence. Discussion Treatment strategies have focused on antifungal activity against Candida species; however, patients continue to suffer from recalcitrant denture-associated stomatitis after antifungal therapy. In this study, we propose a novel approach for managing denture-related candidiasis. Instead of focusing on fungicidal activity, we focused on inhibiting the initial adhesion of Candida. Prevention of adhesion offers a rational method for the long-term control of Candida infection on denture surfaces.1,2 The main feature contributing to the novelty of this study is our mechanism-guided approach to target selection. Two colonization-relevant targets Als3 and GS, were selected to represent different stages of prosthetic surface establishment. Als3 for initial adhesion and GS for early biofilm-support and matrix-associated persistence. In choosing targets relevant to adhesion, our computational pipeline was geared towards clinical observations of denture stomatitis, which is not caused by rampant overgrowth of Candida beneath the denture surface but rather by continuous attachment to the surface. Previous studies in computational AMPs design have often assessed binding to targets with little to no pathogenetic relevance as well as many targets with no clear ranking.5, 6, 7,15,21 Among the three peptides tested, tachystatin showed the most promise for blocking adhesion compared to magainin and the control peptide, as measured by docking scores, molecular dynamics interactions, and binding free energies. This was observed not only because tachystatin showed more favourable comparative energetic trends, but also because it engaged more extensively to conserved regions of the adhesion interface on both Als3 and GS. Further analysis using molecular dynamics showed that Tachystatin immobilizes surface loops and decreases the flexibility of local residues involved in adhesion. Additional evidence for this hypothesis is provided by energetic calculations. MM/PBSA showed that Tachystatin had the lowest and most favourable binding energies overall and van der Waals forces comprised the greatest energetic contributions. This is important mechanistically because a high adhesive force does not necessarily correlate with a strong electrostatic interaction; instead, it can arise from simply sticking to a groove long enough to sterically hinder protein–protein or protein–surface interactions. Peptides that only form electrostatic bonds may exhibit good docking scores but lack sustained activity.8,16,18 However, Magainin only possessed intermediate abilities to interfere with adhesion, with decreased persistence of interaction. Although Magainin has a known mechanism of disrupting membranes, these findings indicate that this mechanism may not be as relevant during prosthetic colonization because initial surface binding occurs before membrane contact can take place. Even the control peptide exhibited poorer surface coverage and lower kinetic stability highlighting that the mechanism of action is critical rather than broad spectrum antifungal properties.9,11 From a translational standpoint, the results of this study may help inform the future development of prosthetic-targeted antifungal strategies. Peptides with predicted biofilm-inhibiting potential such as tachystatin, could be explored in denture bases, surface coatings or localized delivery systems after appropriate experimental validation. If sustained surface exposure can be achieved, adhesion-blocking peptides may help reduce early fungal colonization while potentially exerting less selective pressure than conventional fungicidal approaches. This requires confirmation through in vitro adhesion assays, biofilm models. It also requires material compatibility testing, and cytotoxicity assessment before clinical relevance can be established.2,9,12,13 Importantly, this study also lays out a framework for AMPs selection that can be generalized beyond denture-related candidiasis. Incorporating docking to act as a structural filter with dynamic validation and energetic persistence analysis shifts the focus from selecting diverse peptides to selecting peptides with functional relevance. The described workflow is usable for biofilm infections on medical devices where surface adherence and colonization dictate the infection process.12,14,16 Figure 6 is intended as a conceptual summary rather than an experimental result. It integrates the mechanistic selection criteria used in this study to illustrate how computational screening outputs may guide future experimental prioritization and prosthetic-oriented antifungal development. Conceptual framework for prioritizing adhesion-focused antimicrobial peptides for dental prosthetic applications. The framework was derived from the comparative computational results of the present study and emphasizes adhesion interface occupation, dynamic interaction stability, and energetic persistence as screening criteria. This schematic is intended to guide future experimental prioritization and prosthetic-oriented antifungal development. Some important considerations must be factored into this assessment. As with all in silico studies, our results are limited by the accuracy of the protein structures and force-field representations that are currently available. While molecular dynamics simulations can account for protein flexibility and solvent effects to some degree, they are still unable to fully mimic conditions found in the mouth, such as salivary flow, enzymatic degradation, or interactions with host tissues and immune defences. Furthermore, we did not assess peptide cytotoxicity or material compatibility in this study, which are necessary steps prior to clinical translation and should be addressed in future studies. The current computational study has limitations related to timescale, sampling depth, and model assumptions. The 100-ns molecular dynamics simulations used here are commonly applied for comparative evaluation of peptide–protein interaction stability. They do not fully capture the broader conformational landscape or long-timescale behaviour expected under oral conditions. Single production trajectories were used for each complex without replicate simulations . The observed interaction persistence should be interpreted as a comparative screening result rather than definitive proof of long-term stability. The results remain dependent on the selected force field and simulation setup which applied consistently across all systems, represent an approximation of real biological behaviour. The membrane-associated biological context of GS was simplified in the present docking and simulation workflow. It should be considered when interpreting its predicted interface behaviour. Factors such as salivary flow, protease activity, host interactions and the complex physicochemical environment of the oral cavity were not modelled. Future work should incorporate replicate simulations, extended sampling, and experimental validation in denture-surface systems.16,18,22 The MM/PBSA calculations should be interpreted as approximate endpoint estimates useful for comparative ranking and not as definitive thermodynamic measurements. Future directions will include experimental confirmation of adhesive inhibition through in vitro Candida adhesion studies on denture base surfaces and determining peptide stability during simulated oral use. Lead peptides can also be structurally modified to potentially increase proteolytic stability and surface retention, such as through cyclization or substitution with non-natural amino acids. Guided by the mechanism of action determined in this study, peptides may also be utilized in combination with biomaterials to create new antifungal prosthetic devices capable of preventing infections. Conclusion We present a rational approach for inhibiting denture-related candidiasis that prevents fungal adherence, as opposed to killing fungal cells. Using a computational platform centred on colonization-relevant targets including Als3 and GS we showed that persistent occupation of selected binding interfaces may represent a feasible strategy for prioritizing adhesion-focused antifungal candidates for denture-related applications. Among the tested peptides tachystatin was prioritized for further study because it showed the most favourable overall profile across docking, molecular dynamics and MM/PBSA-based comparative analyses. The study showed promising docking, displayed extended interactions and had positive energy trends within computational limitations. Instead of offering a clinically validated lead candidate this study serves as an exploratory foundation for experimental investigations targeting adhesion-based antifungal approaches and may assist in identifying peptide candidates for denture surfaces. Availability of data and materials The data is available to the corresponding author upon request. Author contributions Conceptualization: Saini, Dermawan. Methodology: Saini, Dermawan. Writing – original draft preparation: Saini. Supervision: Saini, Heboyan. Funding acquisition: Saini. Project administration: Saini, Heboyan. Data curation: Dermawan. Formal analysis: Dermawan, Mosaddad, Alarcón-Sánchez, Heboyan. Writing – review and editing: Dermawan, Mosaddad, Alarcón-Sánchez. Investigation and resource: Dermawan, Chopra, Binduhayyim. Validation: Mosaddad, Alarcón-Sánchez, Heboyan. Funding The authors extend their appreciation to the Deanship of Research and Graduate Studies at King Khalid University for funding this work through Large Research Project under grant number RGP2/405/47. Conflict of interest None disclosed. Acknowledgements All the authors thank King Khalid University, Saudi Arabia, for the financial support. Footnotes Supplementary material associated with this article can be found in the online version at doi:10.1016/j.identj.2026.109693. Contributor Information Seyed Ali Mosaddad, Email: Semosadd@ucm.es. Artak Heboyan, Email: heboyan.artak@gmail.com. Supplementary materials Supplementary material: All the authors thank King Khalid University, Saudi Arabia, for the financial Support. REFERENCES Associated Data Supplementary Materials Supplementary material: All the authors thank King Khalid University, Saudi Arabia, for the financial Support. Data Availability Statement The data is available to the corresponding author upon request.