The Journal of Physical Chemistry Letters

Intrinsic tryptophan fluorescence is widely used as a sensitive reporter of protein conformational dynamics, yet the molecular origin of its temperature-dependent modulation remains unclear. Here we investigate the conformational dynamics of Trp134 in bovine serum albumin (BSA) using molecular dynamics (MD) simulations, free-energy calculations based on umbrella sampling and WHAM, quantum mechanical (QM) calculations, and QM/MM approaches. MD simulations show that the global structure of BSA remains stable while temperature induces a gradual population shift from the Ia+ to the Ia- rotamer. The corresponding free-energy landscapes reveal that this shift arises from subtle changes in basin stability and transition barriers along the rotameric coordinate. In contrast, standalone QM calculations on isolated tryptophan predict different energetic trends, highlighting the sensitivity of rotamer stability to electronic-structure treatments and environmental effects. QM/MM calculations partially reconcile these differences by incorporating the protein environment. Together, these results suggest that temperature reshapes the rotamer free-energy landscape of Trp134, leading to population shifts that modulate intrinsic tryptophan fluorescence in proteins.

The repair of photo-induced DNA lesions through nucleotide excision repair machinery is still the source of important questions. It has been observed that the repair rate of the different cyclobutane pyrimidine dimers, i.e. the photoproducts induced by dimerization of two {pi}-stacked pyrimidines (T<>T, T<>C, C<>T, C<>C), depends on the nucleobases involved in the lesion. TT derivatives (T<>T) are removed more slowly than those containing cytosine, especially in 5. Using all-atom molecular dynamics simulations and free-energy calculations, we demonstrate that the variation of the repair rate observed in human skin and in cultured cutaneous cell is associated to the recognition of the four lesions by the DDB2 protein moiety, and more specifically by the differential structural deformation induced on the complementary strand. Indeed, while C<>C and C<>T induce a larger deviation on the groove parameters, T<>T and T<>C, instead, affect DNA structure to a lesser extent. less affected. These effects then hamper differentially the downstream recruitment of the repair complexes. The observed DNA deformation correlates with the experimental repair rate and provides a structural rationale for the different repair rates of CPD by nucleotide excision repair machinery. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=105 SRC="FIGDIR/small/724087v1_ufig1.gif" ALT="Figure 1"> View larger version (43K): org.highwire.dtl.DTLVardef@cf6b6dorg.highwire.dtl.DTLVardef@195e35forg.highwire.dtl.DTLVardef@1829296org.highwire.dtl.DTLVardef@165baba_HPS_FORMAT_FIGEXP M_FIG C_FIG

The molten globule (MG) state is an intermediate in the unfolding pathway of proteins, typically triggered by denaturing agents such as urea, extreme pH, high pressure, or heat. The microscopic details of such states are far from understood. Here we study the MG states in protein Hen Egg-White Lysozyme (PDB ID: 1AKI) using microscopic constant pH molecular dynamics (CpHMD) simulations and experiments across a wide pH range. We observe that the titratable residues act as key drivers of conformational fluctuations, promoting the emergence of MG states at extreme pH. These states display partial unfolding, and small global structural changes (< 7% deviation). Hydration around the fluctuating acidic residues shows reduced water density and weakened hydrogen bonding at low pH. At high pH, hydration around acidic residues increases relative to pH = 7, whereas hydration around basic residues decreases. The translational and rotational dynamics of hydration water also exhibit pronounced pH dependence: the translational diffusion coefficient (Dtrans) increases linearly with decrease in pH in acidic medium and increases linearly with increasing pH in the basic regime. The rotational diffusion (Drot) shows similar dependencies on pH except a break at pH {approx} 4 corresponding to acidic residue pKa values. Our results may be useful to identify ligand binding of lysozyme in extreme pH conditions.

G-quadruplex (GQ) structures within the HIV-1 long terminal repeat (LTR) regulate viral transcription and represent promising antiviral targets; however, detailed mechanistic understanding of their ligand recognition at the molecular level remains limited and has largely been investigated under dilute conditions despite the crowded and compartmentalized nature of intracellular environment. Here, we investigate the interaction of the cationic porphyrin TMPyP4 with the HIV-1 LTR-III GQ under dilute conditions and inside protein-rich phase-separated condensates that mimic intracellular biocondensates. Steady-state and time-resolved fluorescence measurements reveal a dual binding behavior that is not discernible from absorption spectroscopy. A high-affinity guanine-rich binding mode leads to efficient fluorescence quenching through electron transfer from ground-state guanine to excited TMPyP4, whereas a weaker non-guanine binding mode gives rise to enhanced and long-lived emission. Nucleotide-specific control experiments validate the origin of these distinct binding environments. Molecular docking and molecular dynamics simulations further support preferential binding of TMPyP4 at the terminal G-quartet together with a secondary binding mode near the quadruplex-duplex junction. Importantly, both TMPyP4 and LTR-III GQ preferentially partition into the condensates, where the hybrid GQ structure, dual binding behavior, and associated excited-state signatures remain preserved despite the crowded and viscous environment. Although a slight reduction in binding affinity is observed inside the condensates, the overall binding mechanism remains largely preserved due to compensatory effects arising from the condensate microenvironment. Overall, this work demonstrates that ligand recognition of viral GQ remains preserved within protein condensates and establishes fluorescence spectroscopy as a sensitive approach for resolving hidden binding heterogeneity in GQ-ligand interactions.

Ultraviolet (UV) radiation induces DNA damage associated with cancer and aging, yet the sequence dependence of UV absorption remains to be investigated. Here, we present a systematic study of the UV absorption spectra of DNA based on all-electron Hartree-Fock calculations. We analyze all possible sequences up to four base pairs, as well as longer randomized sequences and genomic nullomers - motifs which are missing in a given genome. We observe a pronounced sequence dependence: cytosine- and guanine-rich motifs exhibit significantly enhanced absorption, whereas adenine-thymine-rich sequences absorb up to four times less in the mid-UV range. Notably, the human genome is biased toward adenine-thymine-rich sequences, giving it an increased susceptibility to UV-induced damage. In addition, we introduce a computational framework enabling spectral calculations of large DNA and RNA fragments, opening the door to large-scale optical analyses.

HIV-1 buds from infected cells as immature virion particles with a scattered envelope glycoprotein (Env) distribution on their envelope. It then undergoes maturation, during which the viral protease cleaves the Gag polyprotein at multiple sites, leading to structural reorganization of the viral particle and lateral redistribution of Env proteins, ultimately rendering the virion infectious. However, the underlying mechanism of maturation-induced Env reorganization remains elusive. In this study, we combine microsecond-long all-atom (AA), bottom-up coarse-grained (CG) molecular dynamics simulations, and diffusion model-based backmapping to investigate the structural organization and key interactions of Env in viral membranes. AA simulations of fully glycosylated Env embedded in HIV-1 mimetic asymmetric bilayers were first performed to characterize its conformational dynamics and Env-lipid interactions. We then developed a bottom-up CG model of glycosylated Env from that AA data and simulated the mature HIV-1 virion envelope containing multiple Env proteins. The CG simulations predict that Env proteins form clusters through interactions mediated by the cytoplasmic tail domain (CTD) and adopt diverse tilted conformations within these clusters. These CG simulations were then backmapped to AA resolution and further AA simulations were carried out to identify, in detail, the specific interacting residues in the Env clusters. Additionally, analysis of epitope accessibility shows that broadly neutralizing antibodies (bnAbs) targeting the V1/V2 and V3 loops may efficiently interact with Env clusters on the mature virion surface. Together, these results provide a molecular mechanism for Env oligomerization during viral maturation and offer new insights into the accessibility of bnAb epitopes on Env clusters.

-Synuclein (Syn) is an intrinsically disordered protein whose interactions with lipid membranes are central to both its physiological function and its role in synucleopathies. While membrane charge, phase, and curvature are each known to influence Syn binding, these properties are typically examined independently, leaving their combined effects on both equilibrium and dynamic membrane association unresolved. Here, we systematically investigate how membrane phase and charge jointly regulate Syn binding, curvature sensitivity, and exchange dynamics using fluorescence microscopy, circular dichroism spectroscopy, and fluorescence recovery after photobleaching (FRAP), complemented by coarse-grained molecular dynamics simulations. Under zwitterionic conditions, Syn preferentially binds highly curved gel-phase membranes, driven by curvature-dependent enrichment of packing defects arising from faceted vesicle morphologies. Incorporation of anionic lipids selectively enhances binding in liquid-phase membranes while attenuating curvature-dependent partitioning in gel-phase membranes. Dynamic measurements reveal that membrane phase and charge also govern the stability of membrane-associated Syn, with gel-phase membranes and anionic lipids promoting kinetically stabilized states. Simulations show that curvature-induced defect formation is strongly amplified in gel-phase membranes but largely insensitive to charge. These findings establish that Syn-membrane interactions are governed by a cooperative interplay between membrane phase, curvature, and charge and highlight the importance of resolving both thermodynamic and kinetic contributions to protein-membrane binding.

Orthoflavivirus, such as West Nile Virus (WNV), dengue virus (DENV) and ZIKA virus (ZIKV), are globally distributed pathogens that pose substantial threats to human health. Currently, there are still no effective antiviral drugs for WNV or ZIKV. Despite the availability of two licensed DENV vaccines, their use remains limited due to potential risks, highlighting an urgent need for antiviral drug development. The highly conserved orthoflavivirus protease NS2B/NS3 is required for viral replication, making it a promising anti-flavivirus target. A major challenge, however, is the highly charged active site of this enzyme, which requires charged chemical matters with low bioavailability. An alternative and more attractive strategy is to target potential allosteric sites or folding intermediate states of the protease. In this work, we employ the topology-based coarse-grained G[o] modeling to explore the coupled binding and folding pathways of WNV NS2B/NS3 protease and study the effects of the widely used experimental construct with a G4SG4 linker between NS2B and NS3 on stability and folding. Our results provide a holistic conformational landscape of the protease binding and folding, including several key intermediate states. We find that the presence of the G4SG4 linker alters the folding pathways and destabilizes the NS2B C-terminus. The latter is consistent with experimental observations that the G4SG4 linked protease has lower activity and adopts an open state without the substrate in crystal structures. Together, these findings provide for the first time a complete picture of the binding and folding of the NS2B/NS3 protease and identify important folding intermediate states that could be targeted for allosteric antiviral drug development. TOC Figure O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=157 SRC="FIGDIR/small/722635v1_ufig1.gif" ALT="Figure 1"> View larger version (40K): org.highwire.dtl.DTLVardef@163c356org.highwire.dtl.DTLVardef@ad7b35org.highwire.dtl.DTLVardef@173ed8aorg.highwire.dtl.DTLVardef@1f026bf_HPS_FORMAT_FIGEXP M_FIG C_FIG

Avian phototransduction and magnetoreception have been proposed to involve shared retinal proteins, including interactions between long-wavelength opsin (LWO), the cone-specific heterotrimeric G protein (Gt), and cryptochrome 4a (Cry4a), yet structural information on avian phototransduction complexes is lacking. Here we present and critically assess two atomistic models of the European robin LWO-Gt complex generated by distinct modelling strategies. A full-complex prediction using AlphaFold3 yields a tightly packed, structurally stable interface but exhibits pronounced activation-like conformational features of the Gt-subunit that persist in simulations of the isolated protein, revealing a strong bias toward the active state. In contrast, a template-guided assembly based on single-chain predictions and an experimental rhodopsin-Gt reference structure forms a weaker interface and shows no intrinsic activation bias, while still displaying subtle activation-related dynamics. These results demonstrate that machine-learned complex prediction can encode functional states independently of the local interaction environment, thereby limiting its interpretability for signalling mechanisms that hinge on activation equilibria. Our findings highlight the need for explicit assessment of conformational-state bias when modelling regulatory protein assemblies and provide a structural framework for future studies of Cry4a-dependent modulation of retinal G-protein signalling in avian magnetoreception.

Sea slugs in the Sacoglossa superorder are some of the few animals capable of photosynthesising by isolating and maintaining functional chloroplasts within their body1,2. While this ability allows some species in this superorder, such as Elysia viridis, to appear green, camouflaging themselves within their surroundings3,4, this species is marked by extremely bright, coloured regions. Here, we show that these animals produce a yet undiscovered class of photonic structure consisting of intracellular mixed amorphous CaCO3 and calcite spherical nanoparticles organised in non-closed-packed face-centred cubic (FCC) lattices and photonic glasses5. By mapping the distribution of the cells containing such architectures, we suggest that their colour is linked both to their function and to their biological formation via the animals renal system. Using a combination of different optical methods and cryo-electron microscopy, we reveal that the biomineralisation pathway proceeds through stages of calcium ion concentration in the kidney, transport via internal vessels, and precipitation from a dense liquid-like precursor, culminating in the formation of monodisperse nanoparticles, which are the building blocks of these photonic structures.

DNA topology is a key regulator of chromatin structure and transcription, yet its direct role in transcription factor recognition remains unclear. Here, we investigate how distinct DNA topological states modulate binding of the Saccharomyces cerevisiae bZIP transcription factor GCN4 using topologically defined plasmids. By combining, complementary biochemical approaches, including Bio-Layer Interferometry applied here for the first time to topology-dependent protein-DNA interactions, we show that DNA supercoiling directly reshapes GCN4-DNA recognition. Positively supercoiled DNA forms more stable and persistent complexes, whereas negatively supercoiled DNA retains greater conformational heterogeneity. To interpret these effects, we performed multiscale molecular simulations. Coarse-grained simulations of plasmids recapitulate the global topology-dependent trends observed experimentally, while matched minicircle models reproduce the same behaviour at the local scale. In strong agreement with experimental data, simulations reveal that DNA topology modulates the conformational ensemble of the GCN4 basic region. Overall, positively supercoiled DNA promotes a more ordered binding mode and localized protein distribution, whereas negatively supercoiled DNA supports increased structural plasticity. These findings identify DNA topology as an active determinant of transcription factor recognition and provide a multiscale framework linking global DNA mechanics to local protein-DNA interactions. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=113 SRC="FIGDIR/small/722604v1_ufig1.gif" ALT="Figure 1"> View larger version (51K): org.highwire.dtl.DTLVardef@18f8ba9org.highwire.dtl.DTLVardef@11a395dorg.highwire.dtl.DTLVardef@ac093borg.highwire.dtl.DTLVardef@923212_HPS_FORMAT_FIGEXP M_FIG C_FIG

The nuclear receptor superfamily is comprised of ligand-regulated transcription factors that contain an intrinsically disordered domain at the amino-terminal end, known as the N-terminal domain (NTD). While this poorly conserved domain is known to possess ligand-independent activation function (AF-1), few NTD functions are conserved between nuclear receptors (NRs). Identified roles in other receptors include androgen receptor (AR), estrogen receptor (ER) and mineralocorticoid receptor (MR). Here, we aim to define the function of the NTD of the farnesoid X receptor (FXR), a crucial regulator of lipid and bile acid metabolism. We show that the NTD engages in interdomain contact with other FXR domains. We also observe that the NTD interacts directly with coregulator proteins. Using mutagenesis, mammalian two-hybrid assays and molecular dynamics simulations, we identify and validate a novel SXXLF motif in the NTD which mediates interactions with both coregulators and the ligand binding domain. Mutation of the motif induces large changes in conformational and allosteric coupling in FXR. Our study identifies a new nuclear receptor-interacting motif that modulates the transcriptional activity of FXR. Graphical AbstractFXR-NTD regulates transcriptional activity through interdomain communication with the LBD and is also involved in co-activator recruitment. The SENLF motif is the first defined functional element within the FXR-NTD and mediates both NTD-LBD interaction and selective co-activator engagements to drive NTD-mediated transcriptional activity. O_FIG O_LINKSMALLFIG WIDTH=135 HEIGHT=200 SRC="FIGDIR/small/724725v1_ufig1.gif" ALT="Figure 1"> View larger version (25K): org.highwire.dtl.DTLVardef@5a37aorg.highwire.dtl.DTLVardef@2fa9e1org.highwire.dtl.DTLVardef@13a19daorg.highwire.dtl.DTLVardef@1775ed2_HPS_FORMAT_FIGEXP M_FIG C_FIG

Designing peptide sequences that remain stable and selective across heterogeneous environments remains a central challenge in biomolecular modeling. Here we introduce an interpretable, physics-based Hamiltonian for environment-conditioned design of -helical peptide sequences. The model integrates helix propensities, pairwise interactions, electrostatics, anisotropic solvent exposure, and interfacial geometry into a unified energy function. To enable comparison across sequence lengths and environments, all contributions are rescaled and expressed as Z-scores relative to random sequence ensembles, yielding a normalized design landscape with balanced physical terms. This formulation defines a structured optimization problem that can be explored using exact, heuristic, and hybrid quantum- classical approaches without modification of the underlying model. The Hamiltonian recovers polar and apolar limits, discriminates experimentally characterized water-soluble and transmembrane -helical peptide sequences, and captures the preferential stabilization of membrane-active sequences at anionic interfaces over non-functional controls. It further enables multi-objective and selective design, generating candidate sequences with tunable environmental specificity.

Curcumin-functionalized gold nanoclusters are promising platforms for catalysis and drug delivery, yet the molecular determinants of their stability, morphology, and solvent response remain unclear. Here, microsecond all-atom molecular dynamics simulations are employed to investigate a 2 nm gold nanoparticle noncovalently coated with different curcumin forms, including neutral enol and trans-keto tautomers, the deprotonated enolate, and their mixtures in water-ethanol and water-methanol solvents. Layer-resolved analyses of radius of gyration, density profiles, and surface coverage reveal that neutral enol and trans forms generate compact assemblies with near-complete surface coverage, whereas enolate-rich systems adopt more expanded conformations with solvent-exposed molecules. Mixed systems preserve these intrinsic packing characteristics while improving overall coverage. Solvent substitution from ethanol to methanol reduces {pi}-{pi} stacking, strengthens Au-curcumin interactions, and increases surface coverage, yielding more compact nanostructures. Free energy and potential of mean force calculations indicate that deprotonated curcumin most effectively screens Au-Au interactions and stabilizes dispersed nanoparticles, while neutral tautomers provide moderate stabilization. Curcumin also enhances the loading of anticancer drug doxorubicin (DOX) onto Au nanoparticles, improving biocompatibility. Enolate(An)-containing systems produce extended structures with weaker membrane interactions, whereas neutral curcumin complexes form compact, positively charged assemblies that strongly bind to negatively charged cancer cell membranes. These findings clarify how tautomeric state and solvent environment cooperatively govern interfacial organization and colloidal stability, establish design guidelines for curcumin-based gold nanocarriers in catalysis, sensing, and drug delivery applications.

The hierarchical organization of multiphase biomolecular condensates into core-shell architectures is a fundamental problem in soft matter and biophysics. While classical explanations rely on hierarchies of interfacial tension ({gamma}) between coexisting liquids, the ultralow tensions of condensates (0.1-1 {micro}N/m) render such hierarchies potentially fragile. We introduce a robust assembly principle based on Polymer-Assisted Condensation (PAC), in which a single polymer species dictates the entire structure. The polymer nucleates a dense core by recruiting a condensation-incompetent protein (P1). A second incompetent protein (P2), which is repelled or otherwise thermodynamically disfavored from entering the polymer-rich core, is nonetheless recruited to the interface by weak attraction to P1, forming a stable shell. This effective repulsion-driven layering operates across a wide parameter space without requiring{gamma} asymmetries and yields a robust structure that is impervious to concentration fluctuations and environmental perturbations. Phase-field modeling and molecular simulations establish this mechanism and capture key features of nucleolar organization. Our work reveals a general physical pathway for encoding spatial order in soft, multicomponent fluids.

Building on a complex between a tubulin protofilament (PF) and a fragment of the Tau protein containing residues 169 to 367, we investigate the dynamics of the disordered elements of the system, namely the tubulin C-terminal tails (CTTs) and the Tau protein, using classical all-atom molecular dynamics simulations. Our results show that CTTs adopt a hook-like dynamic pattern on the bare PF while remaining highly mobile. The binding of Tau on the PF surface alters the dynamics of the I-CTTs in a sequence-dependent manner. While the repeat domains of Tau are mostly maintained on the PF by weak and strong binding patches with the tubulin cores, the Proline-Rich Region (PRR) relies on the wrapping phenomenon of I-CTTs to fuzzily stabilize its interaction with the PF. Our study thus provides a deep dive into the dynamic interplay between the Tau protein and the CTTs of microtubules, the latter being characterized extensively using a variety of disorder-adapted metrics. TOC Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=111 SRC="FIGDIR/small/721901v1_ufig1.gif" ALT="Figure 1"> View larger version (25K): org.highwire.dtl.DTLVardef@b3f985org.highwire.dtl.DTLVardef@1c2bf70org.highwire.dtl.DTLVardef@a66b95org.highwire.dtl.DTLVardef@1e138e0_HPS_FORMAT_FIGEXP M_FIG C_FIG

Reversibly photoswitchable fluorophores are widely used in advanced bioimaging but their design remains demanding. Here, we introduce a new series spanning the whole visible range, which results from combining a large set of fluorogens with the FAST protein scaffold. We first demonstrate that these well-established labeling fluorescent protein tags turn into negative reversible photoswitchers upon decreasing the fluorogen concentration and increasing light intensity. We then show that using not anymore one but two fluorogens adds new responses to illumination. Thus, we obtain positive reversible photoswitchers, that increase their brightness under illumination. We also generate a palette of non-covalent reversibly photoconvertible fluorescent proteins changing their fluorescence color upon illumination, a reversible behavior that still remains absent in regular fluorescent proteins. This light-induced color change opens the possibility to discriminate six spectrally similar FAST variants in live cells upon demonstrating the superiority of using multiple spectral channels for exploiting the time dependence of the fluorescence response to illumination.

The isolation of cellulose nanofibrils (CNFs), a promising precursor for sustainable and high-performance materials, has relied on chemically intensive, energy-demanding processes. As these processes were originally designed for the isolation of CNFs from wood, we herein show that the intrinsic ultrastructure of non-structural plant cells provides unique opportunities, namely direct access to loosely organized cellulose nanonetworks. We demonstrate that this loose nanofibrillar tissue can be transformed into CNFs with sizes down to elementary nanofibrils ([~]4 nm) at high yields (reaching [~]32%) under exceptionally mild hydrothermal conditions. Three distinct plants were evaluated and the physicochemical properties of the obtained nanonetworks and corresponding CNFs were thoroughly studied, including the hydrodynamics of the resulting gels. Films prepared from the obtained CNFs showed similar performance to those obtained from conventionally isolated wood-based CNFs. Overall, this study demonstrates that CNFs can be obtained through low-intensity, hazard-free, processes from widely available biomass. Thus, this approach offers a unique shift in the range of opportunities to produce CNFs facilitating the integration of their use into the food supply chain, biomedical applications, and other regulatory-constrained applications.

Agriphotovoltaics (APV) combines crop production with solar energy generation to address increasing demands for food and energy while reducing land-use competition. Unlike conventional opaque photovoltaic systems, semitransparent organic photovoltaics (OPVs) selectively absorb light, potentially improving efficiency but also altering both light quantity and spectral quality, key factors affecting plant growth. Here, we developed a rapid bioassay based on hypocotyl elongation to evaluate plant responses to OPV-filtered light using Arabidopsis thaliana and Cardamine hirsuta, two species with contrasting shade strategies. Screening a diverse set of OPV materials revealed that plant growth responses depend more on spectral composition than on total light intensity alone. Certain materials, such as PTB7-Th and D18, produced growth patterns similar to neutral shading, while others promoted elongation. Our analyses identified blue light wavelengths, linked to cryptochrome activity, as more critical than red light wavelengths, linked to phytochrome activity, for maintaining normal development. These findings provide a scalable framework to assess OPV-plant compatibility and demonstrate that optimizing spectral quality alongside light intensity is essential for designing efficient APV systems that sustain crop performance while generating renewable energy.

Virtual screening has become an indispensable tool in modern structure-based drug discovery, enabling the identification of candidate molecules by computationally evaluating their potential to bind target proteins. The accuracy of such screenings critically depends on the quality of the target structures employed. Recent advances in protein structure prediction, particularly AlphaFold2, have revolutionized this field with unprecedented accuracy. However, AlphaFold2 models often exhibit limitations in local structural details, especially within binding pockets, which limit their utility for small molecule docking. In contrast, molecular dynamics simulations with accurate atomistic force fields can refine protein structures, but lack the ability to leverage the structural information provided by deep learning approaches. Here, we introduce bAIes, an integrative method that bridges this gap by combining physics-based force fields with data-driven predictions through Bayesian inference. Crucially, bAIes demonstrates a superior ability to discriminate between binders and non-binders in virtual screening campaigns, outperforming both AlphaFold2 and molecular dynamics-refined models. By enhancing the usability of AlphaFold2 models without requiring extensive experimental or computational resources, bAIes offers a convenient solution to a longstanding challenge in structure-based drug design, potentially accelerating the early phases of drug discovery.