Computational Discovery of Antifungal Candidates Against Moniliophthora perniciosa: Multi-target Virtual Screening, Molecular Docking, and QSAR Modeling

Computational Discovery of Antifungal Candidates Against Moniliophthora perniciosa: Multi-target Virtual Screening, Molecular Docking, and QSAR Modeling | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Computational Discovery of Antifungal Candidates Against Moniliophthora perniciosa: Multi-target Virtual Screening, Molecular Docking, and QSAR Modeling Luiz Ricardo Mantovani da Silva This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9391098/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Moniliophthora perniciosa, the causal agent of witches' broom disease of cacao (Theobroma cacao L.), causes annual losses exceeding 30% of Brazilian production, yet no specific fungicide has been developed for this pathosystem. We present the first multi-target computational pipeline for antifungal discovery against M. perniciosa, integrating three fungal-specific targets: lanosterol 14α-demethylase (ERG11/CYP51), alternative oxidase (MpAOX), and chitin synthase class III (CHS). Bioactivity data for 1,398 compounds from ChEMBL were integrated with homology-based structural models and AlphaFold2 predictions. Molecular docking with AutoDock Vina 1.2.5 across five protein structures yielded 92 binding poses. QSAR models (Random Forest, ECFP4 fingerprints) achieved external validation R² of 0.771 (ERG11) and 0.462 (MpAOX). ADMET profiling identified 293 drug-like candidates from 459 pre-filtered compounds. Composite multi-criteria scoring ranked CHEMBL133046 (a farnesyl-hydroquinone derivative; docking affinity −10.16 kcal/mol; IC₅₀ = 2.0 nM; QED = 0.692) as top candidate against MpAOX, alongside N-alkylamide propionamide scaffolds with sub-nanomolar potency against CHS. The results provide a computationally validated shortlist of structurally diverse antifungal candidates with favorable pharmacological profiles for prioritized experimental validation. Computational Biology witches' broom disease Theobroma cacao computer-aided drug design ADMET profiling alternative oxidase chitin synthase Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Introduction Witches’ broom disease, caused by the hemibiotrophic basidiomycete Moniliophthora perniciosa (Stahel) Aime & Phillips-Mora, is the primary biotic constraint on cacao ( Theobroma cacao L.) production in the Americas (Aime and Phillips-Mora 2005 ). The disease provoked a historic collapse of Brazilian cacao output in the 1990s and continues to cause productivity losses of 30–40% in affected plantations, with annual economic impacts exceeding R $ 500 million (Almeida and Albuquerque 2002 ; Brasil 2024 ). Despite decades of research, disease management relies predominantly on labor-intensive phytosanitary pruning and copper-based fungicides whose efficacy is limited and whose sustained use raises environmental concerns (Krauss and Soberanis 2002 ). The chemical control of M. perniciosa is hampered by the absence of fungicides specifically developed for this pathosystem. Most registered antifungals target pathways conserved in agriculturally or clinically relevant fungi such as Aspergillus spp. and Candida spp., without prior evaluation against M. perniciosa. The growing resistance of fungal pathogens to current antifungals, combined with their ecotoxicological profiles, underscores the need for new chemical entities acting on fungal-specific molecular targets. Computational drug discovery — encompassing target identification, virtual screening, molecular docking, and quantitative structure–activity relationship (QSAR) modeling — has become an established strategy for identifying lead compounds in pre-clinical research, substantially reducing the cost and time of early-stage drug discovery (Lionta et al. 2014 ). Despite its widespread application in human medicine and animal health, in silico approaches for plant fungicide discovery, and specifically for M. perniciosa , remain underexplored (Santos et al. 2023 ). Three molecular targets have been prioritized in this study based on their biological essentiality in fungi and the availability of structural homologs: (1) lanosterol 14α-demethylase (ERG11/CYP51), a key enzyme in ergosterol biosynthesis and the molecular target of clinical azoles; (2) alternative oxidase (MpAOX), a mitochondrial terminal oxidase present in M. perniciosa but absent in mammals; and (3) chitin synthase class III (CHS), essential for fungal cell wall integrity and without mammalian ortholog (Meinhardt et al. 2008 ; Loguercio et al. 2009 ). The present study describes a multi-target in silico pipeline integrating structural bioinformatics, virtual screening, molecular docking, and QSAR modeling to identify antifungal candidates against M. perniciosa . This represents the first comprehensive computational antifungal discovery effort directed at this pathosystem. Material and Methods Data acquisition and dataset assembly Bioactivity data for each target were retrieved from the ChEMBL database (version 34) using the chembl-webresource-client Python library. Queries were performed for: (i) ERG11/CYP51 inhibitors (CYP51A homologs in Aspergillus fumigatus and Candida albicans ); (ii) alternative oxidase inhibitors (using Trypanosoma brucei TAO as a structural and functional proxy for MpAOX, given the established mechanistic conservation); and (iii) chitin synthase class III inhibitors from fungal systems. Activity records were filtered to retain only IC₅₀ values (pChEMBL ≥ 5.0, equivalent to IC₅₀ ≤ 10 µM), compounds with valid SMILES, and complete structural information. After duplicate removal and compound expansion using structurally related analogs, the final dataset comprised 1,398 compounds: ERG11 (n = 385), MpAOX (n = 981), and CHS (n = 32). Molecular descriptors and fingerprints were computed with RDKit (version 2023.9.1). ECFP4 Morgan fingerprints (radius = 2, 2048 bits) were used for QSAR modeling. Physicochemical properties including molecular weight (MW), octanol–water partition coefficient (LogP), topological polar surface area (TPSA), quantitative estimate of drug-likeness (QED), and ESOL aqueous solubility (logS) were computed for each compound. Protein structure preparation Three-dimensional structures for molecular docking were obtained by two approaches. For ERG11, a CYP51A homolog structure from Aspergillus fumigatus (PDB: 5TZ1) was used as proxy; for MpAOX, the Trypanosoma brucei alternative oxidase structure (PDB: 3W54) was adopted; for the chitin synthase proxy, the Candida albicans CHS2 structure (PDB: 7STN) was used. Complementarily, AlphaFold2 structure predictions were retrieved from the AlphaFold Protein Structure Database for MpAOX (AF-A8QJP8) and for the M. perniciosa chitin synthase class III domain (MopChs; AF-C0LT25, fragment 249 aa). Receptor structures were prepared with AutoDockTools 1.5.7: water molecules were removed, polar hydrogens were added, and Gasteiger charges were assigned. Ligands were converted to PDBQT format using Open Babel 3.1.1 with energy minimization (MMFF94 force field). Molecular docking Molecular docking was performed with AutoDock Vina 1.2.5 (Eberhardt et al. 2021 ). Docking search boxes were defined by centering on the known active site of each template structure using AutoDockTools. Grid dimensions were set to 22.5 × 22.5 × 22.5 Å, with a spacing of 0.375 Å. Exhaustiveness was set to 8 for initial screening and 16 for top-ranked compounds. A total of 92 docking calculations were performed across five receptor structures. Binding affinity values are reported in kcal/mol; more negative values indicate stronger predicted binding. Poses were visually inspected and ranked by best affinity. QSAR modeling and validation QSAR models were constructed for each target using a Random Forest (RF) algorithm (scikit-learn, 500 estimators). Compound datasets were split into training (80%) and external test (20%) sets using stratified sampling based on pIC₅₀ quartiles. Model performance was evaluated by five-fold cross-validation on the training set (CV-R²) and by external test set metrics including R², Q²F1, RMSE, and MAE. Applicability domain (AD) was estimated using a distance-based approach. Feature importance analysis was performed to identify the most predictive structural features. For ERG11 (n = 385): training set n = 308; external test set n = 77. For MpAOX (n = 981): training set n = 785; external test set n = 196. For CHS (n = 32): dataset size precluded robust external validation; five-fold CV only. ADMET profiling Drug-likeness and ADMET (absorption, distribution, metabolism, excretion, toxicity) profiling was performed for all 459 pre-filtered compounds. Lipinski’s Rule of Five (MW ≤ 500 Da, LogP ≤ 5, H-bond donors ≤ 5, H-bond acceptors ≤ 10), Veber rules (rotatable bonds ≤ 10, TPSA ≤ 140 Ų), and PAINS (Pan-Assay Interference Compounds) filters were applied using RDKit. Blood-brain barrier permeability was estimated via TPSA thresholding (TPSA 90 Ų = low). Aqueous solubility was estimated with the ESOL model. Composite candidate scoring and ranking A composite scoring function was developed to integrate multi-dimensional evidence for each compound. The score incorporated: docking affinity (normalized, weight 30%), experimental pIC₅₀ from ChEMBL (weight 25%), QSAR-predicted pIC₅₀ (weight 20%), QED (weight 15%), and PAINS alert penalty (− 10 per alert). Candidates were ranked by composite score; the top 20 were subjected to scaffold diversity analysis. Results Dataset characterization and molecular properties The assembled dataset comprised 1,398 bioactive compounds across three targets, with pIC₅₀ values ranging from 5.0 to 9.85 (ERG11 median: 6.4; MpAOX median: 6.2; CHS median: 8.3). Molecular property distributions showed that the majority of compounds across all three targets fell within Lipinski-compliant space (Fig. 1 ). ERG11 compounds displayed the highest median MW (432 Da) consistent with the structural complexity of azole derivatives, while CHS inhibitors presented lower MW (median 324 Da) and LogP (median 4.5). Median QED was 0.51 (ERG11), 0.63 (MpAOX), and 0.71 (CHS), indicating overall acceptable drug-likeness. ADMET profiling Of 459 compounds subjected to ADMET profiling, 293 (63.8%) passed all drug-likeness filters and were classified as drug-like (Table 1 ). By target: ERG11 — 236/385 (61.3%), MpAOX — 41/42 screened (97.6%), CHS — 16/32 (50.0%). No compound in the top-20 shortlist triggered PAINS alerts. Blood-brain barrier permeability was estimated as high for six candidates (CHS scaffolds, TPSA < 60 Ų) and medium for the remainder. Aqueous solubility (ESOL logS) ranged from − 3.95 to − 6.02, within acceptable limits for agricultural use. The distribution of compounds in chemical property space is illustrated in Fig. 6 . Table 1 Summary of ADMET profiling results by target. Target Total compounds Drug-like (n) Drug-like (%) Median QED Median logS ERG11 385 236 61.3 0.51 −4.81 MpAOX 42 (screened) 41 97.6 0.63 −5.49 CHS 32 16 50.0 0.71 −4.66 QSAR models QSAR model performance metrics are summarized in Table 2 . In-sample fitting for ERG11 (CV-R² = 0.710 ± 0.041) and CHS (CV-R² = −0.645) is illustrated in Fig. 7 . The ERG11 model demonstrated robust external predictive performance (Fig. 8 ): R²_ext = 0.771, Q²F1 = 0.771 — well above the accepted threshold (Q²F1 > 0.60; Roy et al. 2015 ) — with RMSE of 0.524 pIC₅₀ units and an overfitting gap of 0.08, indicating good generalization. The MpAOX model (Fig. 9 ) achieved CV-R² = 0.413 ± 0.046 and external R² = 0.462 (RMSE = 0.724), acceptable given the broad chemical diversity of the training set and the inter-species proxy nature of the bioactivity data. The CHS model (n = 32; CV-R² = −0.645) was underpowered; CHS candidate ranking therefore relied primarily on experimental pIC₅₀ and docking affinity. Table 2 QSAR model performance metrics for each target. Target n (total) n (train) n (test) CV-R² (mean ± SD) R² (ext) Q²F1 RMSE (ext) MAE (ext) Overfit gap ERG11 385 308 77 0.710 ± 0.041 0.771 0.771 0.524 0.418 0.08 (minimal) MpAOX 981 785 196 0.413 ± 0.046 0.462 0.462 0.724 0.573 0.29 (moderate) CHS 32 — — −0.645 ± — — — 1.363 — — (underpowered) Molecular docking Molecular docking was performed for the top-ranked ligands against five receptor structures: MpAOX proxy (3W54), ERG11 proxy (5TZ1), CHS proxy (7STN), MpAOX AlphaFold (AF-A8QJP8), and MopChs AlphaFold (AF-C0LT25). A total of 92 binding poses were generated (Table 3 summarizes results by receptor). Best docking affinities ranged from − 7.3 kcal/mol (CHS proxy) to − 10.16 kcal/mol (MpAOX proxy, CHEMBL133046). The MpAOX binding site yielded the strongest predicted affinities, consistent with the architectural depth of the AOX ubiquinol-binding cavity. Pairwise correlation between docking affinity and experimental pChEMBL is shown in Fig. 2 ; it was moderate for ERG11 (r = 0.61, n = 20), weak and negative for MpAOX (r = − 0.22, n = 13), and negligible for CHS (r = 0.03, n = 19), reflecting inherent limitations of proxy-based docking where receptor conformations correspond to heterologous species. Boxplot analysis (Fig. 3 ) confirmed that the − 9.0 kcal/mol threshold distinguished a distinct high-affinity subset (11 compounds), all of which appeared in the final top-20 ranking. Table 3 Summary of molecular docking results by receptor structure. Receptor Structure n poses Best affinity (kcal/mol) Top compound MpAOX proxy 3W54 (TAO T. brucei ) 32 −10.16 CHEMBL133046 ERG11 proxy 5TZ1 (CYP51A A. fumigatus ) 20 −8.84 CHEMBL808 CHS proxy 7STN (CHS2 C. albicans ) 20 −7.89 CHEMBL265094 MpAOX AlphaFold AF-A8QJP8 20 −8.81 CHEMBL36446 MopChs AlphaFold AF-C0LT25 20 −8.10 CHEMBL35735 Top antifungal candidates The integrated composite scoring ranked 20 candidates; Table 4 presents the top 10. The leading compound, CHEMBL133046 — a terpenoid phenol derivative (MW 420.9 Da, LogP 5.22, QED 0.692) — achieved the highest docking affinity against MpAOX (− 10.16 kcal/mol), with an experimental pIC₅₀ of 8.70 (IC₅₀ = 2.0 nM against TAO proxy) and QSAR-predicted pIC₅₀ of 7.56. It satisfied Lipinski and Veber criteria, bore no PAINS alerts, and showed medium BBB permeability (TPSA 83.8 Ų), favorable for foliar application with limited systemic accumulation. Compounds CHEMBL4291644 and CHEMBL4294528 — structurally related farnesyl-hydroquinone derivatives (MW 319–323 Da, LogP ~ 5.1, QED ≥ 0.807) — ranked second and third, with docking affinities of − 9.10 and − 9.19 kcal/mol respectively against MpAOX and experimental IC₅₀ values of 3.98 nM. Their high QED scores (≥ 0.807) and low TPSA (57.5–64.2 Ų) indicate superior drug-likeness. Among CHS-targeted candidates, CHEMBL36446 (MW 310.4 Da, LogP 4.14, QED 0.729) showed the highest experimental activity (IC₅₀ = 0.9 nM; pChEMBL = 9.05) and docking affinity of − 7.89 kcal/mol. CHEMBL36409 presented the most potent experimental IC₅₀ (0.14 nM; pChEMBL = 9.85) with QSAR-predicted pIC₅₀ of 7.95. Notably, CHS inhibitors in this study belong to the class of N-alkylamide propionamide derivatives, a scaffold previously unexplored in the M. perniciosa context. Table 4 Top 10 antifungal candidates ranked by composite score. Rank Compound Target Docking (kcal/mol) IC₅₀ exp (nM) QSAR pIC₅₀ QED MW (Da) LogP Score 1 CHEMBL133046 MpAOX −10.16 2.0 7.56 0.692 420.9 5.22 60.9 2 CHEMBL4291644 MpAOX −9.10 3.98 7.44 0.842 319.8 5.17 59.0 3 CHEMBL4294528 MpAOX −9.19 3.98 7.49 0.807 322.8 5.11 58.7 4 CHEMBL36446 CHS −7.89 0.9 8.00 0.729 310.4 4.14 57.2 5 CHEMBL36240 CHS −7.61 1.1 8.33 0.734 352.5 5.31 57.0 6 CHEMBL35735 CHS −8.10 1.23 7.95 0.680 324.5 4.53 56.3 7 CHEMBL36409 CHS −7.30 0.14 7.95 0.737 294.4 3.20 55.5 8 CHEMBL36445 CHS −7.48 0.68 8.23 0.665 324.5 4.53 55.2 9 CHEMBL4741395 CHS −7.50 ~ 1.0 7.75 0.769 296.4 2.65 54.8 10 CHEMBL808 ERG11 −8.84 ~ 5.0 6.88 0.610 381.7 3.94 53.1 Chemical diversity Scaffold analysis of the top 20 candidates (Fig. 10 ) revealed exclusively heterocyclic scaffolds, distributed across three structural series: farnesyl-hydroquinone derivatives (MpAOX, ranks 1–3), N-alkylamide propionamides (CHS, ranks 4–9), and azole/sterol analogues (ERG11, rank 10). Tanimoto similarity to clinical azoles (fluconazole, itraconazole, voriconazole) was consistently low (< 0.25), indicating structural novelty relative to registered antifungals. Discussion This study presents the first multi-target computational antifungal pipeline for Moniliophthora perniciosa , identifying structurally diverse candidates acting on three essential fungal targets. The results demonstrate that in silico approaches can generate actionable hypotheses for experimental validation in this poorly characterized pathosystem. Target selection rationale. ERG11/CYP51 is the molecular target of all clinical triazoles and has been validated as an antifungal target in multiple fungal species; its presence in M. perniciosa is well established (Meinhardt et al. 2008 ). MpAOX represents a particularly attractive target because AOX is absent in mammals and plants and has been identified as a virulence factor in several plant-pathogenic fungi (Teixeira et al. 2015 ). CHS class III participates in chitin deposition in the primary cell septum and is essential for hyphal growth; its fungal specificity makes it an ideal selectivity anchor. The use of proxy structures from well-characterized organisms was a deliberate strategic choice to circumvent the scarcity of M. perniciosa co-crystal structures, a limitation shared by most emerging plant fungal pathogens. QSAR model quality. The ERG11 QSAR model achieved external R² = 0.771, exceeding the threshold of 0.60 commonly required for regulatory-grade QSAR (OECD 2007 ). The MpAOX model (external R² = 0.462) is acceptable given the broad chemical diversity of the training set and the inter-species bioactivity transfer; however, predictions for MpAOX candidates should be interpreted with appropriate uncertainty. The CHS model failure (CV-R² = −0.645) directly reflects the small dataset size (n = 32), and CHS candidate ranking relied primarily on docking affinity and experimental pIC₅₀. Expanding the CHS dataset through additional ChEMBL queries and integration of NuBBE (natural products database) is recommended as a priority for model improvement. Lead candidates and antifungal potential. CHEMBL133046, a farnesyl-hydroquinone derivative, is particularly noteworthy. This compound class is structurally related to terpenoid natural products produced by cacao endophytes, raising the possibility of synergistic biological and chemical control strategies. Its IC₅₀ of 2.0 nM against the AOX proxy and docking affinity of − 10.16 kcal/mol position it among the strongest computational candidates identified to date for any M. perniciosa target. N-alkylamide propionamide compounds (CHEMBL36446, CHEMBL36409) show exceptionally high potency against CHS (IC₅₀ < 1 nM), a scaffold family not previously reported in phytopathological contexts. Limitations. Several limitations must be acknowledged. (i) All bioactivity data derive from heterologous assay systems (primarily Candida , Aspergillus , and T. brucei ); direct M. perniciosa activity remains to be experimentally determined. (ii) Proxy receptor structures introduce systematic docking errors relative to the true M. perniciosa binding site geometry. (iii) AlphaFold models, while structurally plausible, have not been experimentally validated for M. perniciosa proteins. (iv) The composite scoring function, although integrated, is heuristic and has not been validated against an independent experimental benchmark in this system. These limitations underscore that the present results should be interpreted as prioritized hypotheses for experimental validation, not as confirmed bioactivity predictions. Implications for disease management. Computational discovery as described here complements the biological control and genetic resistance strategies reviewed in the witches’ broom management literature (Loguercio et al. 2009 ; Pomella et al. 2008 ; Fister et al. 2018 ). Fungicide-based control of M. perniciosa has historically been limited by the lack of selective compounds; the identification of structurally novel candidates targeting MpAOX and CHS — both absent in the cacao host — opens a rational pathway for developing selective fungicides. Integration of computational-guided fungicide discovery with biocontrol and resistant variety programs represents an emerging and promising horizon for integrated witches’ broom management. Conclusion This study established a validated multi-target in silico pipeline for antifungal discovery against Moniliophthora perniciosa and identified ten structurally diverse, drug-like candidates with strong predicted affinity for ERG11, MpAOX, and CHS. The ERG11 QSAR model achieved regulatory-grade external predictivity (Q²F1 = 0.771), and molecular docking identified CHEMBL133046 as the top candidate with predicted binding affinity of − 10.16 kcal/mol. These results provide a computationally validated shortlist for prioritized experimental testing — including in vitro MIC determination against M. perniciosa isolates, target engagement assays, and selectivity evaluation — constituting a first step toward a fungicide discovery program tailored to the cacao–witches’ broom pathosystem. Declarations Acknowledgements The author thanks the Centro Universitário do Sagrado Coração (UNISAGRADO) for institutional support. Computational analyses were performed using open-source bioinformatics and cheminformatics tools (AutoDock Vina, RDKit, Open Babel, scikit-learn). Protein structures were obtained from the RCSB Protein Data Bank and the AlphaFold Protein Structure Database. Author Contributions Luiz Ricardo Mantovani da Silva: Conceptualization, Methodology, Software, Formal analysis, Data curation, Investigation, Visualization, Writing – original draft, Writing – review & editing. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data Availability All data, code, and pipeline configurations are available at the project repository. Docking input and output files, QSAR models (.pkl), ADMET results, and figures are included in the supplementary data package. Additional information may be requested from the corresponding author. Competing interests: The author declares no competing interests. Ethics approval: Not applicable. AI assistance: AI-assisted tools were used for code generation and language editing; the author takes full responsibility for all content, scientific interpretations, and conclusions. References Aime MC, Phillips-Mora W (2005) The causal agents of witches’ broom and frosty pod rot of cacao (chocolate, Theobroma cacao ) form a new lineage of Marasmiaceae. Mycologia 97:1012–1022. https://doi.org/10.3852/mycologia.97.5.1012 Almeida LHCC, Albuquerque PSB (2002) Epidemiologia e danos econômicos da vassoura de bruxa do cacaueiro. In: Albuquerque PSB, Luz EDMN, Pires JL (eds) A vassoura de bruxa do cacaueiro: patossistema, avanços científicos e tecnológicos e perspectivas de controle. CEPLAC/CEPEC, Ilhéus, pp 131–153 Brasil (2024) Produção de cacau no Brasil — estatísticas. Ministério da Agricultura, Pecuária e Abastecimento (MAPA). Accessed January 2025. https://www.gov.br/agricultura/pt-br/assuntos/producao-animal/cacau Eberhardt J, Santos-Martins D, Tillack AF, Forli S (2021) AutoDock Vina 1.2.0: new docking methods, expanded force field, and python bindings. Journal of Chemical Information and Modeling 61:3891–3898. https://doi.org/10.1021/acs.jcim.1c00203 Fister AS, Landherr L, Maximova SN, Guiltinan MJ (2018) Transient expression of CRISPR/Cas9 machinery targeting TcNPR3 enhances defense response in Theobroma cacao . Frontiers in Plant Science 9:268. https://doi.org/10.3389/fpls.2018.00268 Krauss U, Soberanis W (2002) Effect of fertilization and biocontrol agent application frequency on cocoa pod health. Biological Control 24:82–89. https://doi.org/10.1016/S1049-9644(02)00014-6 Lionta E, Spyrou G, Vassilatis DK, Cournia Z (2014) Structure-based virtual screening for drug discovery: principles, applications and recent advances. Current Topics in Medicinal Chemistry 14:1923–1938. https://doi.org/10.2174/1568026614666140929124445 Loguercio LL, Lins LMS, de Almeida ACB, Oliveira JS, Pires ABL (2009) Selection of Trichoderma stromaticum isolates for efficient biological control of witches’ broom disease in cacao. Biological Control 51:130–139. https://doi.org/10.1016/j.biocontrol.2009.05.012 Meinhardt LW, Rincones J, Bailey BA, Aime MC, Griffith GW, Zhang D, Pereira GAG (2008) Moniliophthora perniciosa , the causal agent of witches’ broom disease of cacao: what’s new from this old foe? Molecular Plant Pathology 9:577–588. https://doi.org/10.1111/j.1364-3703.2008.00490.x OECD (2007) Guidance document on the validation of (quantitative) structure-activity relationship [(Q)SAR] models. OECD Series on Testing and Assessment No. 69. OECD Publishing, Paris Pomella AWV, Bailey BA, Bae H, Melnick RL, de Macedo FE, Thomaziello RA, Niella GR, Samuels GJ, Hebbar PK (2008) Trichoderma stromaticum for management of witches’ broom of cacao in Brazil. Biological Control 46:414–425. https://doi.org/10.1016/j.biocontrol.2008.04.023 Roy K, Kar S, Ambure P (2015) On a simple approach for determining applicability domain of QSAR models. Chemometrics and Intelligent Laboratory Systems 145:22–29. https://doi.org/10.1016/j.chemolab.2015.04.013 Santos AS, Santos FA, Oliveira JM, Rodrigues LA, Lima FMS, de Figueiredo NE, Figueira A, Cascardo JCM, Barroso-Carvalho ML (2023) State of the Art of the Molecular Biology of the Interaction between Cocoa and Witches’ Broom Disease: A Systematic Review. International Journal of Molecular Sciences 24:5684. https://doi.org/10.3390/ijms24065684 Teixeira PJPL, Thomazella DPT, Pereira GAG (2015) Time for cocoa: an integrated approach to characterize the molecular mechanisms of the Theobroma cacao – Moniliophthora perniciosa pathosystem. Phytopathology 105:936–946. https://doi.org/10.1094/PHYTO-11-14-0326-IA Additional Declarations The authors declare no competing interests. Supplementary Files admetresults.csv Admet_results Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9391098","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":621566634,"identity":"27a5226f-1d3a-4b7e-bd76-a40357bc611f","order_by":0,"name":"Luiz Ricardo Mantovani da Silva","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4UlEQVRIiWNgGAWjYDADNv7mgw+ANA8fEYoZG8BaJI4lG4C0sBGthYEhR00CrJeQet32s8cffGyzy+NjOMNW+TXHToaNgfnhoxt4tJidyUtsnNmWXMzG3Hvstuy2ZKDD2IyNc/BpOZBj2MxzhjmxjeFc2m3JbcxALTxs0ni1nH8D0lIP1JJjViy5rZ4ILTdAtlQcBmth/LjtMDFa3hjOnFFxPLENGMjSjNuO87AxE/LL+RyDDx8MqhPn9zcf/PhzW7U9P3vzw8f4tKAAZh4wSaxyEGD8QYrqUTAKRsEoGDEAAMsuSA/G9Le+AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-3053-8306","institution":"Universidade do Sagrado Coração - UNISAGRADO","correspondingAuthor":true,"prefix":"","firstName":"Luiz","middleName":"Ricardo Mantovani da","lastName":"Silva","suffix":""}],"badges":[],"createdAt":"2026-04-12 02:07:42","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9391098/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9391098/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106877919,"identity":"fc4e4465-e4ad-4403-8fc8-4e52078ff88c","added_by":"auto","created_at":"2026-04-14 10:42:51","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2094927,"visible":true,"origin":"","legend":"\u003cp\u003eMolecular property distributions across the three target datasets (ERG11, MpAOX, CHS). Histograms show frequency distributions of molecular weight (MW, Da), octanol–water partition coefficient (LogP), and quantitative estimate of drug-likeness (QED). Dashed vertical lines indicate Lipinski Rule of Five thresholds (MW = 500 Da; LogP = 5).\u003c/p\u003e","description":"","filename":"Fig1propertydistributions.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/e09dbf0f831a5720aafd2e49.png"},{"id":106961892,"identity":"cc0e46bf-6a49-470b-9da0-b9f11541c3bf","added_by":"auto","created_at":"2026-04-15 09:27:36","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2624640,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation between molecular docking affinity (kcal/mol) and experimental bioactivity (pChEMBL, −log IC₅₀) for compounds docked against each proxy receptor: MpAOX proxy (PDB: 3W54; r = −0.22, n = 13), ERG11 proxy (PDB: 5TZ1; r = 0.61, n = 20), and CHS proxy (PDB: 7STN; r = 0.03, n = 19). Dashed lines indicate linear regression fits; top-ranked compounds are annotated by ChEMBL identifier.\u003c/p\u003e","description":"","filename":"Fig2dockingvspchembl.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/c02cd3bcb42a92cf45277a10.png"},{"id":106960700,"identity":"d78b13d0-85a7-4407-9a22-20ae7da99d88","added_by":"auto","created_at":"2026-04-15 09:22:43","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":979151,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of molecular docking binding affinities (kcal/mol) across five receptor structures: MpAOX proxy (3W54), ERG11 proxy (5TZ1), CHS proxy (7STN), MpAOX AlphaFold model (AF-A8QJP8), and MopChs AlphaFold model (AF-C0LT25). Box plots show median, interquartile range, and individual data points. The horizontal dashed line at −9.0 kcal/mol indicates the high-affinity threshold used for candidate prioritization.\u003c/p\u003e","description":"","filename":"Fig3dockingboxplot.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/357c032f3cca486a2627f2ec.png"},{"id":106877927,"identity":"40e109c5-160c-4699-ba50-bf036b5f270f","added_by":"auto","created_at":"2026-04-14 10:42:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2007612,"visible":true,"origin":"","legend":"\u003cp\u003eComparative heatmap of normalized pharmacological profiles for the top 10 antifungal candidates. Each metric (docking affinity, experimental pIC₅₀, QED, molecular weight, LogP) was normalized to a 0–1 scale; green indicates favorable values and red indicates unfavorable values. Rows represent individual compounds; columns represent scoring dimensions.\u003c/p\u003e","description":"","filename":"Fig4candidatesheatmap.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/05b13e2613840e1f1ce2a59b.png"},{"id":106877924,"identity":"09c1e0d7-5442-4697-a1cc-f63bae0f4964","added_by":"auto","created_at":"2026-04-14 10:42:51","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":4654563,"visible":true,"origin":"","legend":"\u003cp\u003eRadar (spider) charts representing the multi-dimensional pharmacological profile of the four highest-ranked antifungal candidates. Each axis represents a normalized metric: docking affinity, QSAR-predicted pIC₅₀, QED, drug-like molecular weight (MW \u0026lt; 500 Da), and optimal LogP range (proximity to LogP = 3). Shaded areas indicate the composite pharmacological profile.\u003c/p\u003e","description":"","filename":"Fig5radartop4.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/51f22cd97acec7f2ae6927f6.png"},{"id":106877928,"identity":"9c89973e-1e1d-4fdd-98d8-595394f42b42","added_by":"auto","created_at":"2026-04-14 10:42:52","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":7389616,"visible":true,"origin":"","legend":"\u003cp\u003eChemical space distribution of ADMET-profiled compounds (n = 459). (Left) Molecular weight (MW) versus octanol–water partition coefficient (LogP) scatter plot; the shaded region denotes Lipinski Rule of Five-compliant space. Drug-like compounds (filled circles) and non-drug-like compounds (crosses) are distinguished by target. (Right) Quantitative estimate of drug-likeness (QED) versus topological polar surface area (TPSA); the dashed threshold at TPSA = 140 Ų indicates the Veber permeability criterion.\u003c/p\u003e","description":"","filename":"Fig6chemicalspace.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/9d2828eb9b8a9ee98795ce2d.png"},{"id":106877925,"identity":"bfc2640c-97dd-4c05-b145-a89f2090d419","added_by":"auto","created_at":"2026-04-14 10:42:51","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":4977908,"visible":true,"origin":"","legend":"\u003cp\u003eIn-sample QSAR model performance for ERG11 (n = 385; CV-R² = 0.710, RMSE = 0.540) and CHS (n = 32; CV-R² = −0.645, RMSE = 1.363). Predicted pIC₅₀ (Random Forest) versus experimental pIC₅₀ is shown for training-set compounds; the dashed diagonal represents the ideal prediction line (y = x).\u003c/p\u003e","description":"","filename":"Fig7qsarpredictions.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/1724897c0d27938e02750213.png"},{"id":106877922,"identity":"9c2e0449-02fb-4c03-bcc7-2fc1451d98d9","added_by":"auto","created_at":"2026-04-14 10:42:51","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1816537,"visible":true,"origin":"","legend":"\u003cp\u003eExternal validation of the ERG11 QSAR model (training set n = 308; test set n = 77). (Left) Predicted versus observed pIC₅₀ for test-set compounds (diamonds) and training-set compounds (circles), alongside the identity line. (Right) Residual plot for test-set predictions. Model metrics: R²_ext = 0.771, Q²F1 = 0.771, RMSE = 0.524 pIC₅₀ units, MAE = 0.418 pIC₅₀ units.\u003c/p\u003e","description":"","filename":"Fig8qsarerg11validation.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/1dd64c04bb33268a8d85aa5e.png"},{"id":106877923,"identity":"75ff2568-e54e-4edf-8539-5dcc29557f84","added_by":"auto","created_at":"2026-04-14 10:42:51","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":2742784,"visible":true,"origin":"","legend":"\u003cp\u003eExternal validation of the MpAOX QSAR model (training set n = 785; test set n = 196). (Left) Predicted versus observed pIC₅₀ for test-set compounds (diamonds) and training-set compounds (circles), alongside the identity line. (Right) Residual plot for test-set predictions. Model metrics: R²_ext = 0.462, Q²F1 = 0.462, RMSE = 0.724 pIC₅₀ units, MAE = 0.573 pIC₅₀ units.\u003c/p\u003e","description":"","filename":"Fig9qsarmpaoxvalidation.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/4ad305fcae79fb198ad57339.png"},{"id":106877926,"identity":"dd8d94d7-ebde-43ec-9ed9-54c57cef134a","added_by":"auto","created_at":"2026-04-14 10:42:51","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":2679655,"visible":true,"origin":"","legend":"\u003cp\u003eScaffold novelty analysis of the top 20 antifungal candidates by composite score. (Left) Distribution of chemical scaffold classes. (Right) Maximum Tanimoto similarity (ECFP4 fingerprints) of each candidate to three reference clinical azoles (fluconazole, voriconazole, itraconazole). All 20 top-ranked candidates fall below the structural novelty threshold of 0.30 (green dashed line), with maximum similarities below 0.15, indicating that none are analogues of registered antifungal drugs.\u003c/p\u003e","description":"","filename":"Fig10scaffolddiversity.png","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/efa4476754894cb3cc567358.png"},{"id":106963472,"identity":"91aa5146-a74a-4fb0-b0bd-43aef4511d63","added_by":"auto","created_at":"2026-04-15 09:44:41","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":28564048,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/56d0bf63-8431-457e-aacb-b53d3ded0c61.pdf"},{"id":106877920,"identity":"a1a1c973-6d28-4ef1-9ab1-ed2a75400969","added_by":"auto","created_at":"2026-04-14 10:42:51","extension":"csv","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":218199,"visible":true,"origin":"","legend":"\u003cp\u003eAdmet_results\u003c/p\u003e","description":"","filename":"admetresults.csv","url":"https://assets-eu.researchsquare.com/files/rs-9391098/v1/717257181eaf69f191d2edc8.csv"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eComputational Discovery of Antifungal Candidates Against Moniliophthora perniciosa:\u003c/p\u003e\n\u003cp\u003eMulti-target Virtual Screening, Molecular Docking, and QSAR Modeling\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWitches\u0026rsquo; broom disease, caused by the hemibiotrophic basidiomycete \u003cem\u003eMoniliophthora perniciosa\u003c/em\u003e (Stahel) Aime \u0026amp; Phillips-Mora, is the primary biotic constraint on cacao (\u003cem\u003eTheobroma cacao\u003c/em\u003e L.) production in the Americas (Aime and Phillips-Mora \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). The disease provoked a historic collapse of Brazilian cacao output in the 1990s and continues to cause productivity losses of 30\u0026ndash;40% in affected plantations, with annual economic impacts exceeding R\u003cspan\u003e$\u003c/span\u003e 500\u0026nbsp;million (Almeida and Albuquerque \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Brasil \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Despite decades of research, disease management relies predominantly on labor-intensive phytosanitary pruning and copper-based fungicides whose efficacy is limited and whose sustained use raises environmental concerns (Krauss and Soberanis \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2002\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe chemical control of M. perniciosa is hampered by the absence of fungicides specifically developed for this pathosystem. Most registered antifungals target pathways conserved in agriculturally or clinically relevant fungi such as Aspergillus spp. and Candida spp., without prior evaluation against M. perniciosa. The growing resistance of fungal pathogens to current antifungals, combined with their ecotoxicological profiles, underscores the need for new chemical entities acting on fungal-specific molecular targets.\u003c/p\u003e \u003cp\u003eComputational drug discovery \u0026mdash; encompassing target identification, virtual screening, molecular docking, and quantitative structure\u0026ndash;activity relationship (QSAR) modeling \u0026mdash; has become an established strategy for identifying lead compounds in pre-clinical research, substantially reducing the cost and time of early-stage drug discovery (Lionta et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Despite its widespread application in human medicine and animal health, \u003cem\u003ein silico\u003c/em\u003e approaches for plant fungicide discovery, and specifically for \u003cem\u003eM. perniciosa\u003c/em\u003e, remain underexplored (Santos et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThree molecular targets have been prioritized in this study based on their biological essentiality in fungi and the availability of structural homologs: (1) lanosterol 14α-demethylase (ERG11/CYP51), a key enzyme in ergosterol biosynthesis and the molecular target of clinical azoles; (2) alternative oxidase (MpAOX), a mitochondrial terminal oxidase present in \u003cem\u003eM. perniciosa\u003c/em\u003e but absent in mammals; and (3) chitin synthase class III (CHS), essential for fungal cell wall integrity and without mammalian ortholog (Meinhardt et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Loguercio et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe present study describes a multi-target \u003cem\u003ein silico\u003c/em\u003e pipeline integrating structural bioinformatics, virtual screening, molecular docking, and QSAR modeling to identify antifungal candidates against \u003cem\u003eM. perniciosa\u003c/em\u003e. This represents the first comprehensive computational antifungal discovery effort directed at this pathosystem.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Material and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData acquisition and dataset assembly\u003c/h2\u003e \u003cp\u003eBioactivity data for each target were retrieved from the ChEMBL database (version 34) using the chembl-webresource-client Python library. Queries were performed for: (i) ERG11/CYP51 inhibitors (CYP51A homologs in \u003cem\u003eAspergillus fumigatus\u003c/em\u003e and \u003cem\u003eCandida albicans\u003c/em\u003e); (ii) alternative oxidase inhibitors (using \u003cem\u003eTrypanosoma brucei\u003c/em\u003e TAO as a structural and functional proxy for MpAOX, given the established mechanistic conservation); and (iii) chitin synthase class III inhibitors from fungal systems. Activity records were filtered to retain only IC₅₀ values (pChEMBL\u0026thinsp;\u0026ge;\u0026thinsp;5.0, equivalent to IC₅₀ \u0026le; 10 \u0026micro;M), compounds with valid SMILES, and complete structural information. After duplicate removal and compound expansion using structurally related analogs, the final dataset comprised 1,398 compounds: ERG11 (n\u0026thinsp;=\u0026thinsp;385), MpAOX (n\u0026thinsp;=\u0026thinsp;981), and CHS (n\u0026thinsp;=\u0026thinsp;32).\u003c/p\u003e \u003cp\u003eMolecular descriptors and fingerprints were computed with RDKit (version 2023.9.1). ECFP4 Morgan fingerprints (radius\u0026thinsp;=\u0026thinsp;2, 2048 bits) were used for QSAR modeling. Physicochemical properties including molecular weight (MW), octanol\u0026ndash;water partition coefficient (LogP), topological polar surface area (TPSA), quantitative estimate of drug-likeness (QED), and ESOL aqueous solubility (logS) were computed for each compound.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eProtein structure preparation\u003c/h3\u003e\n\u003cp\u003eThree-dimensional structures for molecular docking were obtained by two approaches. For ERG11, a CYP51A homolog structure from \u003cem\u003eAspergillus fumigatus\u003c/em\u003e (PDB: 5TZ1) was used as proxy; for MpAOX, the \u003cem\u003eTrypanosoma brucei\u003c/em\u003e alternative oxidase structure (PDB: 3W54) was adopted; for the chitin synthase proxy, the \u003cem\u003eCandida albicans\u003c/em\u003e CHS2 structure (PDB: 7STN) was used. Complementarily, AlphaFold2 structure predictions were retrieved from the AlphaFold Protein Structure Database for MpAOX (AF-A8QJP8) and for the \u003cem\u003eM. perniciosa\u003c/em\u003e chitin synthase class III domain (MopChs; AF-C0LT25, fragment 249 aa). Receptor structures were prepared with AutoDockTools 1.5.7: water molecules were removed, polar hydrogens were added, and Gasteiger charges were assigned. Ligands were converted to PDBQT format using Open Babel 3.1.1 with energy minimization (MMFF94 force field).\u003c/p\u003e\n\u003ch3\u003eMolecular docking\u003c/h3\u003e\n\u003cp\u003eMolecular docking was performed with AutoDock Vina 1.2.5 (Eberhardt et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Docking search boxes were defined by centering on the known active site of each template structure using AutoDockTools. Grid dimensions were set to 22.5 \u0026times; 22.5 \u0026times; 22.5 \u0026Aring;, with a spacing of 0.375 \u0026Aring;. Exhaustiveness was set to 8 for initial screening and 16 for top-ranked compounds. A total of 92 docking calculations were performed across five receptor structures. Binding affinity values are reported in kcal/mol; more negative values indicate stronger predicted binding. Poses were visually inspected and ranked by best affinity.\u003c/p\u003e\n\u003ch3\u003eQSAR modeling and validation\u003c/h3\u003e\n\u003cp\u003eQSAR models were constructed for each target using a Random Forest (RF) algorithm (scikit-learn, 500 estimators). Compound datasets were split into training (80%) and external test (20%) sets using stratified sampling based on pIC₅₀ quartiles. Model performance was evaluated by five-fold cross-validation on the training set (CV-R\u0026sup2;) and by external test set metrics including R\u0026sup2;, Q\u0026sup2;F1, RMSE, and MAE. Applicability domain (AD) was estimated using a distance-based approach. Feature importance analysis was performed to identify the most predictive structural features.\u003c/p\u003e \u003cp\u003eFor ERG11 (n\u0026thinsp;=\u0026thinsp;385): training set n\u0026thinsp;=\u0026thinsp;308; external test set n\u0026thinsp;=\u0026thinsp;77.\u003c/p\u003e \u003cp\u003eFor MpAOX (n\u0026thinsp;=\u0026thinsp;981): training set n\u0026thinsp;=\u0026thinsp;785; external test set n\u0026thinsp;=\u0026thinsp;196.\u003c/p\u003e \u003cp\u003eFor CHS (n\u0026thinsp;=\u0026thinsp;32): dataset size precluded robust external validation; five-fold CV only.\u003c/p\u003e\n\u003ch3\u003eADMET profiling\u003c/h3\u003e\n\u003cp\u003eDrug-likeness and ADMET (absorption, distribution, metabolism, excretion, toxicity) profiling was performed for all 459 pre-filtered compounds. Lipinski\u0026rsquo;s Rule of Five (MW\u0026thinsp;\u0026le;\u0026thinsp;500 Da, LogP\u0026thinsp;\u0026le;\u0026thinsp;5, H-bond donors\u0026thinsp;\u0026le;\u0026thinsp;5, H-bond acceptors\u0026thinsp;\u0026le;\u0026thinsp;10), Veber rules (rotatable bonds\u0026thinsp;\u0026le;\u0026thinsp;10, TPSA\u0026thinsp;\u0026le;\u0026thinsp;140 \u0026Aring;\u0026sup2;), and PAINS (Pan-Assay Interference Compounds) filters were applied using RDKit. Blood-brain barrier permeability was estimated via TPSA thresholding (TPSA\u0026thinsp;\u0026lt;\u0026thinsp;60 \u0026Aring;\u0026sup2; = high; 60\u0026ndash;90 \u0026Aring;\u0026sup2; = medium; \u0026gt; 90 \u0026Aring;\u0026sup2; = low). Aqueous solubility was estimated with the ESOL model.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eComposite candidate scoring and ranking\u003c/h2\u003e \u003cp\u003eA composite scoring function was developed to integrate multi-dimensional evidence for each compound. The score incorporated: docking affinity (normalized, weight 30%), experimental pIC₅₀ from ChEMBL (weight 25%), QSAR-predicted pIC₅₀ (weight 20%), QED (weight 15%), and PAINS alert penalty (\u0026minus;\u0026thinsp;10 per alert). Candidates were ranked by composite score; the top 20 were subjected to scaffold diversity analysis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eDataset characterization and molecular properties\u003c/h2\u003e \u003cp\u003eThe assembled dataset comprised 1,398 bioactive compounds across three targets, with pIC₅₀ values ranging from 5.0 to 9.85 (ERG11 median: 6.4; MpAOX median: 6.2; CHS median: 8.3). Molecular property distributions showed that the majority of compounds across all three targets fell within Lipinski-compliant space (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). ERG11 compounds displayed the highest median MW (432 Da) consistent with the structural complexity of azole derivatives, while CHS inhibitors presented lower MW (median 324 Da) and LogP (median 4.5). Median QED was 0.51 (ERG11), 0.63 (MpAOX), and 0.71 (CHS), indicating overall acceptable drug-likeness.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eADMET profiling\u003c/h2\u003e \u003cp\u003eOf 459 compounds subjected to ADMET profiling, 293 (63.8%) passed all drug-likeness filters and were classified as drug-like (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). By target: ERG11 \u0026mdash; 236/385 (61.3%), MpAOX \u0026mdash; 41/42 screened (97.6%), CHS \u0026mdash; 16/32 (50.0%). No compound in the top-20 shortlist triggered PAINS alerts. Blood-brain barrier permeability was estimated as high for six candidates (CHS scaffolds, TPSA\u0026thinsp;\u0026lt;\u0026thinsp;60 \u0026Aring;\u0026sup2;) and medium for the remainder. Aqueous solubility (ESOL logS) ranged from \u0026minus;\u0026thinsp;3.95 to \u0026minus;\u0026thinsp;6.02, within acceptable limits for agricultural use. The distribution of compounds in chemical property space is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of ADMET profiling results by target.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTarget\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal compounds\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDrug-like (n)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDrug-like (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMedian QED\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMedian logS\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eERG11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e385\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e236\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e61.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;4.81\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMpAOX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e42 (screened)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e97.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;5.49\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e50.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;4.66\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eQSAR models\u003c/h2\u003e \u003cp\u003eQSAR model performance metrics are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. In-sample fitting for ERG11 (CV-R\u0026sup2; = 0.710\u0026thinsp;\u0026plusmn;\u0026thinsp;0.041) and CHS (CV-R\u0026sup2; = \u0026minus;0.645) is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e. The ERG11 model demonstrated robust external predictive performance (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e): R\u0026sup2;_ext\u0026thinsp;=\u0026thinsp;0.771, Q\u0026sup2;F1\u0026thinsp;=\u0026thinsp;0.771 \u0026mdash; well above the accepted threshold (Q\u0026sup2;F1\u0026thinsp;\u0026gt;\u0026thinsp;0.60; Roy et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) \u0026mdash; with RMSE of 0.524 pIC₅₀ units and an overfitting gap of 0.08, indicating good generalization. The MpAOX model (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e) achieved CV-R\u0026sup2; = 0.413\u0026thinsp;\u0026plusmn;\u0026thinsp;0.046 and external R\u0026sup2; = 0.462 (RMSE\u0026thinsp;=\u0026thinsp;0.724), acceptable given the broad chemical diversity of the training set and the inter-species proxy nature of the bioactivity data. The CHS model (n\u0026thinsp;=\u0026thinsp;32; CV-R\u0026sup2; = \u0026minus;0.645) was underpowered; CHS candidate ranking therefore relied primarily on experimental pIC₅₀ and docking affinity.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eQSAR model performance metrics for each target.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTarget\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003en (total)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003en (train)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003en (test)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCV-R\u0026sup2; (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eR\u0026sup2; (ext)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eQ\u0026sup2;F1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eRMSE (ext)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eMAE (ext)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eOverfit gap\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eERG11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e385\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e308\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.710\u0026thinsp;\u0026plusmn;\u0026thinsp;0.041\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.771\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.771\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.524\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.418\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.08 (minimal)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMpAOX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e981\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e785\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e196\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.413\u0026thinsp;\u0026plusmn;\u0026thinsp;0.046\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.462\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.462\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.724\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.573\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.29 (moderate)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.645 \u0026plusmn; \u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u0026mdash; (underpowered)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eMolecular docking\u003c/h2\u003e \u003cp\u003eMolecular docking was performed for the top-ranked ligands against five receptor structures: MpAOX proxy (3W54), ERG11 proxy (5TZ1), CHS proxy (7STN), MpAOX AlphaFold (AF-A8QJP8), and MopChs AlphaFold (AF-C0LT25). A total of 92 binding poses were generated (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e summarizes results by receptor).\u003c/p\u003e \u003cp\u003eBest docking affinities ranged from \u0026minus;\u0026thinsp;7.3 kcal/mol (CHS proxy) to \u0026minus;\u0026thinsp;10.16 kcal/mol (MpAOX proxy, CHEMBL133046). The MpAOX binding site yielded the strongest predicted affinities, consistent with the architectural depth of the AOX ubiquinol-binding cavity. Pairwise correlation between docking affinity and experimental pChEMBL is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e; it was moderate for ERG11 (r\u0026thinsp;=\u0026thinsp;0.61, n\u0026thinsp;=\u0026thinsp;20), weak and negative for MpAOX (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.22, n\u0026thinsp;=\u0026thinsp;13), and negligible for CHS (r\u0026thinsp;=\u0026thinsp;0.03, n\u0026thinsp;=\u0026thinsp;19), reflecting inherent limitations of proxy-based docking where receptor conformations correspond to heterologous species. Boxplot analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) confirmed that the \u0026minus;\u0026thinsp;9.0 kcal/mol threshold distinguished a distinct high-affinity subset (11 compounds), all of which appeared in the final top-20 ranking.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of molecular docking results by receptor structure.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eReceptor\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStructure\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003en poses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBest affinity (kcal/mol)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTop compound\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMpAOX proxy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3W54 (TAO \u003cem\u003eT. brucei\u003c/em\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;10.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCHEMBL133046\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eERG11 proxy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5TZ1 (CYP51A \u003cem\u003eA. fumigatus\u003c/em\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;8.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCHEMBL808\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCHS proxy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7STN (CHS2 \u003cem\u003eC. albicans\u003c/em\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;7.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCHEMBL265094\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMpAOX AlphaFold\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAF-A8QJP8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;8.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCHEMBL36446\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMopChs AlphaFold\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAF-C0LT25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;8.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCHEMBL35735\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eTop antifungal candidates\u003c/h2\u003e \u003cp\u003eThe integrated composite scoring ranked 20 candidates; Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e presents the top 10. The leading compound, CHEMBL133046 \u0026mdash; a terpenoid phenol derivative (MW 420.9 Da, LogP 5.22, QED 0.692) \u0026mdash; achieved the highest docking affinity against MpAOX (\u0026minus;\u0026thinsp;10.16 kcal/mol), with an experimental pIC₅₀ of 8.70 (IC₅₀ = 2.0 nM against TAO proxy) and QSAR-predicted pIC₅₀ of 7.56. It satisfied Lipinski and Veber criteria, bore no PAINS alerts, and showed medium BBB permeability (TPSA 83.8 \u0026Aring;\u0026sup2;), favorable for foliar application with limited systemic accumulation.\u003c/p\u003e \u003cp\u003eCompounds CHEMBL4291644 and CHEMBL4294528 \u0026mdash; structurally related farnesyl-hydroquinone derivatives (MW 319\u0026ndash;323 Da, LogP\u0026thinsp;~\u0026thinsp;5.1, QED\u0026thinsp;\u0026ge;\u0026thinsp;0.807) \u0026mdash; ranked second and third, with docking affinities of \u0026minus;\u0026thinsp;9.10 and \u0026minus;\u0026thinsp;9.19 kcal/mol respectively against MpAOX and experimental IC₅₀ values of 3.98 nM. Their high QED scores (\u0026ge;\u0026thinsp;0.807) and low TPSA (57.5\u0026ndash;64.2 \u0026Aring;\u0026sup2;) indicate superior drug-likeness.\u003c/p\u003e \u003cp\u003eAmong CHS-targeted candidates, CHEMBL36446 (MW 310.4 Da, LogP 4.14, QED 0.729) showed the highest experimental activity (IC₅₀ = 0.9 nM; pChEMBL\u0026thinsp;=\u0026thinsp;9.05) and docking affinity of \u0026minus;\u0026thinsp;7.89 kcal/mol. CHEMBL36409 presented the most potent experimental IC₅₀ (0.14 nM; pChEMBL\u0026thinsp;=\u0026thinsp;9.85) with QSAR-predicted pIC₅₀ of 7.95. Notably, CHS inhibitors in this study belong to the class of N-alkylamide propionamide derivatives, a scaffold previously unexplored in the \u003cem\u003eM. perniciosa\u003c/em\u003e context.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTop 10 antifungal candidates ranked by composite score.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRank\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCompound\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTarget\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDocking (kcal/mol)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eIC₅₀ exp (nM)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eQSAR pIC₅₀\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eQED\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eMW (Da)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eLogP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eScore\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL133046\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMpAOX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;10.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.692\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e420.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e5.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e60.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL4291644\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMpAOX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;9.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.842\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e319.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e5.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e59.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL4294528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMpAOX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;9.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.807\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e322.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e5.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e58.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL36446\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;7.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e8.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.729\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e310.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e57.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL36240\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;7.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e8.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.734\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e352.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e5.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e57.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL35735\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;8.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.680\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e324.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e56.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL36409\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;7.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.737\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e294.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e55.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL36445\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;7.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e8.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.665\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e324.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e55.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL4741395\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCHS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;7.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e~\u0026thinsp;1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.769\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e296.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e54.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCHEMBL808\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eERG11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;8.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e~\u0026thinsp;5.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e6.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.610\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e381.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e53.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eChemical diversity\u003c/h2\u003e \u003cp\u003eScaffold analysis of the top 20 candidates (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e) revealed exclusively heterocyclic scaffolds, distributed across three structural series: farnesyl-hydroquinone derivatives (MpAOX, ranks 1\u0026ndash;3), N-alkylamide propionamides (CHS, ranks 4\u0026ndash;9), and azole/sterol analogues (ERG11, rank 10). Tanimoto similarity to clinical azoles (fluconazole, itraconazole, voriconazole) was consistently low (\u0026lt;\u0026thinsp;0.25), indicating structural novelty relative to registered antifungals.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study presents the first multi-target computational antifungal pipeline for \u003cem\u003eMoniliophthora perniciosa\u003c/em\u003e, identifying structurally diverse candidates acting on three essential fungal targets. The results demonstrate that \u003cem\u003ein silico\u003c/em\u003e approaches can generate actionable hypotheses for experimental validation in this poorly characterized pathosystem.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTarget selection rationale.\u003c/b\u003e ERG11/CYP51 is the molecular target of all clinical triazoles and has been validated as an antifungal target in multiple fungal species; its presence in \u003cem\u003eM. perniciosa\u003c/em\u003e is well established (Meinhardt et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). MpAOX represents a particularly attractive target because AOX is absent in mammals and plants and has been identified as a virulence factor in several plant-pathogenic fungi (Teixeira et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). CHS class III participates in chitin deposition in the primary cell septum and is essential for hyphal growth; its fungal specificity makes it an ideal selectivity anchor. The use of proxy structures from well-characterized organisms was a deliberate strategic choice to circumvent the scarcity of \u003cem\u003eM. perniciosa\u003c/em\u003e co-crystal structures, a limitation shared by most emerging plant fungal pathogens.\u003c/p\u003e \u003cp\u003e \u003cb\u003eQSAR model quality.\u003c/b\u003e The ERG11 QSAR model achieved external R\u0026sup2; = 0.771, exceeding the threshold of 0.60 commonly required for regulatory-grade QSAR (OECD \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). The MpAOX model (external R\u0026sup2; = 0.462) is acceptable given the broad chemical diversity of the training set and the inter-species bioactivity transfer; however, predictions for MpAOX candidates should be interpreted with appropriate uncertainty. The CHS model failure (CV-R\u0026sup2; = \u0026minus;0.645) directly reflects the small dataset size (n\u0026thinsp;=\u0026thinsp;32), and CHS candidate ranking relied primarily on docking affinity and experimental pIC₅₀. Expanding the CHS dataset through additional ChEMBL queries and integration of NuBBE (natural products database) is recommended as a priority for model improvement.\u003c/p\u003e \u003cp\u003e \u003cb\u003eLead candidates and antifungal potential.\u003c/b\u003e CHEMBL133046, a farnesyl-hydroquinone derivative, is particularly noteworthy. This compound class is structurally related to terpenoid natural products produced by cacao endophytes, raising the possibility of synergistic biological and chemical control strategies. Its IC₅₀ of 2.0 nM against the AOX proxy and docking affinity of \u0026minus;\u0026thinsp;10.16 kcal/mol position it among the strongest computational candidates identified to date for any \u003cem\u003eM. perniciosa\u003c/em\u003e target. N-alkylamide propionamide compounds (CHEMBL36446, CHEMBL36409) show exceptionally high potency against CHS (IC₅₀ \u0026lt; 1 nM), a scaffold family not previously reported in phytopathological contexts.\u003c/p\u003e \u003cp\u003e \u003cb\u003eLimitations.\u003c/b\u003e Several limitations must be acknowledged. (i) All bioactivity data derive from heterologous assay systems (primarily \u003cem\u003eCandida\u003c/em\u003e, \u003cem\u003eAspergillus\u003c/em\u003e, and \u003cem\u003eT. brucei\u003c/em\u003e); direct \u003cem\u003eM. perniciosa\u003c/em\u003e activity remains to be experimentally determined. (ii) Proxy receptor structures introduce systematic docking errors relative to the true \u003cem\u003eM. perniciosa\u003c/em\u003e binding site geometry. (iii) AlphaFold models, while structurally plausible, have not been experimentally validated for \u003cem\u003eM. perniciosa\u003c/em\u003e proteins. (iv) The composite scoring function, although integrated, is heuristic and has not been validated against an independent experimental benchmark in this system. These limitations underscore that the present results should be interpreted as prioritized hypotheses for experimental validation, not as confirmed bioactivity predictions.\u003c/p\u003e \u003cp\u003e \u003cb\u003eImplications for disease management.\u003c/b\u003e Computational discovery as described here complements the biological control and genetic resistance strategies reviewed in the witches\u0026rsquo; broom management literature (Loguercio et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Pomella et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Fister et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Fungicide-based control of \u003cem\u003eM. perniciosa\u003c/em\u003e has historically been limited by the lack of selective compounds; the identification of structurally novel candidates targeting MpAOX and CHS \u0026mdash; both absent in the cacao host \u0026mdash; opens a rational pathway for developing selective fungicides. Integration of computational-guided fungicide discovery with biocontrol and resistant variety programs represents an emerging and promising horizon for integrated witches\u0026rsquo; broom management.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study established a validated multi-target \u003cem\u003ein silico\u003c/em\u003e pipeline for antifungal discovery against \u003cem\u003eMoniliophthora perniciosa\u003c/em\u003e and identified ten structurally diverse, drug-like candidates with strong predicted affinity for ERG11, MpAOX, and CHS. The ERG11 QSAR model achieved regulatory-grade external predictivity (Q\u0026sup2;F1\u0026thinsp;=\u0026thinsp;0.771), and molecular docking identified CHEMBL133046 as the top candidate with predicted binding affinity of \u0026minus;\u0026thinsp;10.16 kcal/mol. These results provide a computationally validated shortlist for prioritized experimental testing \u0026mdash; including \u003cem\u003ein vitro\u003c/em\u003e MIC determination against \u003cem\u003eM. perniciosa\u003c/em\u003e isolates, target engagement assays, and selectivity evaluation \u0026mdash; constituting a first step toward a fungicide discovery program tailored to the cacao\u0026ndash;witches\u0026rsquo; broom pathosystem.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eThe author thanks the Centro Universit\u0026aacute;rio do Sagrado Cora\u0026ccedil;\u0026atilde;o (UNISAGRADO) for institutional support. Computational analyses were performed using open-source bioinformatics and cheminformatics tools (AutoDock Vina, RDKit, Open Babel, scikit-learn). Protein structures were obtained from the RCSB Protein Data Bank and the AlphaFold Protein Structure Database.\u003c/p\u003e\n\u003cp\u003eAuthor Contributions\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLuiz Ricardo Mantovani da Silva:\u003c/strong\u003e Conceptualization, Methodology, Software, Formal analysis, Data curation, Investigation, Visualization, Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003eData Availability\u003c/p\u003e\n\u003cp\u003eAll data, code, and pipeline configurations are available at the project repository. Docking input and output files, QSAR models (.pkl), ADMET results, and figures are included in the supplementary data package. Additional information may be requested from the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests:\u003c/strong\u003e The author declares no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAI assistance:\u003c/strong\u003e AI-assisted tools were used for code generation and language editing; the author takes full responsibility for all content, scientific interpretations, and conclusions.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAime MC, Phillips-Mora W (2005) The causal agents of witches\u0026rsquo; broom and frosty pod rot of cacao (chocolate, \u003cem\u003eTheobroma cacao\u003c/em\u003e) form a new lineage of Marasmiaceae.\u0026nbsp;\u003cem\u003eMycologia\u003c/em\u003e 97:1012\u0026ndash;1022. https://doi.org/10.3852/mycologia.97.5.1012\u003c/li\u003e\n \u003cli\u003eAlmeida LHCC, Albuquerque PSB (2002) Epidemiologia e danos econ\u0026ocirc;micos da vassoura de bruxa do cacaueiro. In: Albuquerque PSB, Luz EDMN, Pires JL (eds) A vassoura de bruxa do cacaueiro: patossistema, avan\u0026ccedil;os cient\u0026iacute;ficos e tecnol\u0026oacute;gicos e perspectivas de controle. CEPLAC/CEPEC, Ilh\u0026eacute;us, pp 131\u0026ndash;153\u003c/li\u003e\n \u003cli\u003eBrasil (2024) Produ\u0026ccedil;\u0026atilde;o de cacau no Brasil \u0026mdash; estat\u0026iacute;sticas. Minist\u0026eacute;rio da Agricultura, Pecu\u0026aacute;ria e Abastecimento (MAPA). Accessed January 2025. https://www.gov.br/agricultura/pt-br/assuntos/producao-animal/cacau\u003c/li\u003e\n \u003cli\u003eEberhardt J, Santos-Martins D, Tillack AF, Forli S (2021) AutoDock Vina 1.2.0: new docking methods, expanded force field, and python bindings. \u003cem\u003eJournal of Chemical Information and Modeling\u003c/em\u003e 61:3891\u0026ndash;3898. https://doi.org/10.1021/acs.jcim.1c00203\u003c/li\u003e\n \u003cli\u003eFister AS, Landherr L, Maximova SN, Guiltinan MJ (2018) Transient expression of CRISPR/Cas9 machinery targeting TcNPR3 enhances defense response in \u003cem\u003eTheobroma cacao\u003c/em\u003e. \u003cem\u003eFrontiers in Plant Science\u003c/em\u003e 9:268. https://doi.org/10.3389/fpls.2018.00268\u003c/li\u003e\n \u003cli\u003eKrauss U, Soberanis W (2002) Effect of fertilization and biocontrol agent application frequency on cocoa pod health. \u003cem\u003eBiological Control\u003c/em\u003e 24:82\u0026ndash;89. https://doi.org/10.1016/S1049-9644(02)00014-6\u003c/li\u003e\n \u003cli\u003eLionta E, Spyrou G, Vassilatis DK, Cournia Z (2014) Structure-based virtual screening for drug discovery: principles, applications and recent advances. \u003cem\u003eCurrent Topics in Medicinal Chemistry\u003c/em\u003e 14:1923\u0026ndash;1938. https://doi.org/10.2174/1568026614666140929124445\u003c/li\u003e\n \u003cli\u003eLoguercio LL, Lins LMS, de Almeida ACB, Oliveira JS, Pires ABL (2009) Selection of \u003cem\u003eTrichoderma stromaticum\u003c/em\u003e isolates for efficient biological control of witches\u0026rsquo; broom disease in cacao. \u003cem\u003eBiological Control\u003c/em\u003e 51:130\u0026ndash;139. https://doi.org/10.1016/j.biocontrol.2009.05.012\u003c/li\u003e\n \u003cli\u003eMeinhardt LW, Rincones J, Bailey BA, Aime MC, Griffith GW, Zhang D, Pereira GAG (2008) \u003cem\u003eMoniliophthora perniciosa\u003c/em\u003e, the causal agent of witches\u0026rsquo; broom disease of cacao: what\u0026rsquo;s new from this old foe? \u003cem\u003eMolecular Plant Pathology\u003c/em\u003e 9:577\u0026ndash;588. https://doi.org/10.1111/j.1364-3703.2008.00490.x\u003c/li\u003e\n \u003cli\u003eOECD (2007) Guidance document on the validation of (quantitative) structure-activity relationship [(Q)SAR] models. OECD Series on Testing and Assessment No.\u0026nbsp;69. OECD Publishing, Paris\u003c/li\u003e\n \u003cli\u003ePomella AWV, Bailey BA, Bae H, Melnick RL, de Macedo FE, Thomaziello RA, Niella GR, Samuels GJ, Hebbar PK (2008) \u003cem\u003eTrichoderma stromaticum\u003c/em\u003e for management of witches\u0026rsquo; broom of cacao in Brazil. \u003cem\u003eBiological Control\u003c/em\u003e 46:414\u0026ndash;425. https://doi.org/10.1016/j.biocontrol.2008.04.023\u003c/li\u003e\n \u003cli\u003eRoy K, Kar S, Ambure P (2015) On a simple approach for determining applicability domain of QSAR models. \u003cem\u003eChemometrics and Intelligent Laboratory Systems\u003c/em\u003e 145:22\u0026ndash;29. https://doi.org/10.1016/j.chemolab.2015.04.013\u003c/li\u003e\n \u003cli\u003eSantos AS, Santos FA, Oliveira JM, Rodrigues LA, Lima FMS, de Figueiredo NE, Figueira A, Cascardo JCM, Barroso-Carvalho ML (2023) State of the Art of the Molecular Biology of the Interaction between Cocoa and Witches\u0026rsquo; Broom Disease: A Systematic Review. \u003cem\u003eInternational Journal of Molecular Sciences\u003c/em\u003e 24:5684. https://doi.org/10.3390/ijms24065684\u003c/li\u003e\n \u003cli\u003eTeixeira PJPL, Thomazella DPT, Pereira GAG (2015) Time for cocoa: an integrated approach to characterize the molecular mechanisms of the \u003cem\u003eTheobroma cacao\u003c/em\u003e\u0026ndash;\u003cem\u003eMoniliophthora perniciosa\u003c/em\u003e pathosystem. \u003cem\u003ePhytopathology\u003c/em\u003e 105:936\u0026ndash;946. https://doi.org/10.1094/PHYTO-11-14-0326-IA\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"UNIVERSIDADE DO SAGRADO CORAÇÃO - UNISAGRADO","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"witches' broom disease, Theobroma cacao, computer-aided drug design, ADMET profiling, alternative oxidase, chitin synthase","lastPublishedDoi":"10.21203/rs.3.rs-9391098/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9391098/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cem\u003eMoniliophthora perniciosa, the causal agent of witches' broom disease of cacao (Theobroma cacao L.), causes annual losses exceeding 30% of Brazilian production, yet no specific fungicide has been developed for this pathosystem. We present the first multi-target computational pipeline for antifungal discovery against M. perniciosa, integrating three fungal-specific targets: lanosterol 14α-demethylase (ERG11/CYP51), alternative oxidase (MpAOX), and chitin synthase class III (CHS). Bioactivity data for 1,398 compounds from ChEMBL were integrated with homology-based structural models and AlphaFold2 predictions. Molecular docking with AutoDock Vina 1.2.5 across five protein structures yielded 92 binding poses. QSAR models (Random Forest, ECFP4 fingerprints) achieved external validation R² of 0.771 (ERG11) and 0.462 (MpAOX). ADMET profiling identified 293 drug-like candidates from 459 pre-filtered compounds. Composite multi-criteria scoring ranked CHEMBL133046 (a farnesyl-hydroquinone derivative; docking affinity −10.16 kcal/mol; IC₅₀ = 2.0 nM; QED = 0.692) as top candidate against MpAOX, alongside N-alkylamide propionamide scaffolds with sub-nanomolar potency against CHS. The results provide a computationally validated shortlist of structurally diverse antifungal candidates with favorable pharmacological profiles for prioritized experimental validation.\u003c/em\u003e\u003c/p\u003e","manuscriptTitle":"Computational Discovery of Antifungal Candidates Against Moniliophthora perniciosa:\nMulti-target Virtual Screening, Molecular Docking, and QSAR Modeling","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-14 10:42:33","doi":"10.21203/rs.3.rs-9391098/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8fc9ea29-130b-4d3b-a332-d43da67e78df","owner":[],"postedDate":"April 14th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":66147173,"name":"Computational Biology"}],"tags":[],"updatedAt":"2026-04-14T10:42:33+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-14 10:42:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9391098","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9391098","identity":"rs-9391098","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}