Abstract: Background: HIV-infected children (CLHIV) are known to have impaired seroconversion or waning immunity following seroconversion against measles. This may result in measles infection, morbidity and mortality among CLHIV and is also a challenge to measles elimination. Objectives: To compare measles seroprevalence between vaccinated CLHIV and HIV-exposed uninfected children (HEU). The null hypothesis (H₀) was that the difference in proportions for measles seroconversion between CLHIV and HEU is not statistically significant. Materials and Methods: An unmatched case-control study was conducted among 35 CLHIV and 28 HEU aged 18 months to 8 years, attending the Paediatric Centre of Excellence in HIV care (PCOE) at Medical College, Kolkata, between 2022-2024. Participants provided demographic and treatment details; venous blood sample was collected and tested using an ELISA kit to detect measles IgG. We measured outcome as distribution of measles IgG titre and odds ratio (OR) for seroconversion with 95% confidence interval (CI). Results: Groups were similar by sex, weight and duration since last measles vaccine, but differed in age. There were 50% of HEU and 33.7% of CLHIV who had seroconverted, with Measles IgG antibody ≥ 275 IU/L. The overall seroconversion rate was 41%; the difference in proportion did not reachstatistical significance (χ² = 1.7; p = 0.18). However, antibody distributions according to serostatus were significantly different by the Mann-Whitney U test (p = 0.04).Conclusion: Null hypothesis could not be rejected; a combination of vaccine failures and waning immunity possibly resulted in low seroconversion. This indicates existing pockets of vulnerability which could hinder measles elimination.

Abstract: Because of its anatomical complexity, the cervical spine is highly susceptible to injury, especially the blunt acceleration/deceleration trauma of which the most frequent mechanism, is “whiplash”, commonly referred to as Whiplash Associated Disorder (WAD). Despite this knowledge the significance and complexity of whiplash injuries are widely underestimated. This underestimation of the significance of this injury is the widely held belief that it is a benign self-limiting soft tissue injury that frequently has monetary gain attached. This negative framing of the injury which has largely been shaped by the insurance industry rather than by clinicians frequently causes medical providers to view the injury with skepticism. Adding to the problem faced by clinicians are the more recently imposed guidelines for diagnosing and treating acute whiplash, especially in the area of imaging that has placed both the provider and the patient at increased risk.

Abstract: The performance of a sequencing batch biofilter granular reactor (SBBGR), followed by a dual media granular activated carbon (GAC) column, was evaluated in terms of its ability to remove selected per- and polyfluoroalkyl substances (PFAS) from landfill leachate. The results show that the SBBGR achieved an overall reduction of 51%, with the preferential removal of long-chain PFAS, while short-chain PFAS were only partially removed. Subsequent GAC treatment exhibited compound-specific breakthrough behaviour, which was governed by chain length. Short-chain PFAS (e.g., perfluorobutanoic acid) exhibited rapid bed volumes at 50% breakthrough (BV₅₀ ≈ 88), whereas long-chain PFAS (e.g., perfluorooctanoic acid and perfluorooctanesulfonic acid) were substantially more retained (BV₅₀ ≈ 446 and 361, respectively), with perfluorohexanesulfonic acid and perfluorodecanoic acid failing to reach BV₅₀ within the monitored period. Mass balance analysis showed that the hybrid GAC column captured ~73% of the influent PFAS mass. This resulted in >80–95% retention of long-chain PFAS and < 40% retention of short-chain PFAS. Although long-chain PFAS were preferentially adsorbed, mobile short-chain species dominated residual effluent loads. These findings highlight the need for optimized contact times or dual-media strategies to control the breakthrough of short-chain PFAS.

Abstract: Background/Objectives: Parvovirus infection cause sever diseases in both feline and ca-nine species, mostly affects adult cats and dogs, but cause higher risk in the kittens and puppies. This virus is known to be contagious; the simple way of spreading occurs through food and shelter as well as hands and clothing of people. The recovered species may continue to shed parvovirus in their feces for an extended period, leading to severe environmental contamination. There is no universal vaccine available that protect dogs and cats against parvovirus infections. The main goal of this study is to design a pan parvovirus multiepitope DNA based vaccine that could be administered to dog and cats. Methods: We utilized AI-machine learning incorporated server tools such as IEDB and NetMHCpan to predict B-cell and T-cell epitopes. VaxiJen and ToxinPred were used to analyze immune characteristic features and docking with feline alleles using HADdock server. Following, the immune response and stability of vaccine construct was confirmed with disulfide engineering, Normal mode analysis and molecular docking, and dynamic simulations were done with Toll like receptors of both feline and canine (TLR4 and TLR5) for 50ns. The triggered immune response was determined with immuno-simulation (ImmSim) and their activity in biological environment was reinforced with in silico cloning. Results: The B cell epitopes (NS1 - 9, NS2 - 4, VP1- 12 and VP2- 9) predicted with IEDB database were subjected to antigenicity prediction. MHC class I and IFN prediction and MHC class II and IL-4 prediction were done with IEDB and NetMHCpan. The T cell epitopes showed high binding affinities with the feline alleles. The final vaccine was de-signed by combining the top-ranked B-cell epitopes T- cell epitopes, filtered from high antigenic, non-allergic, non-toxic and good solubility, and with the better binding affinity score of the structural and non- structural proteins (NS1, NS2, VP1 and VP2) of feline and canine parvoviruses through linkers and adjuvants. The disulfide bond prediction and Normal mode analysis showed our vaccine construct are stable and flexible. The molecular docking analysis was performed between the designed vaccine epitopes and the TLRs (TLR4 – feline and TLR5 – canine) with Biovia Discovery Studio using Zdocker, it showed the better binding interaction with value of 22.26 (Zdock score), -47.409 (Zrank score) for feline and 16.54 (Zdock score), -134.295 (Zrank score) for canine. Conclusions: The pan multi-epitope DNA based vaccine combining the four major structural proteins (NS1, NS2, VP1 and VP2) possess dual purpose to protect both feline and canine species against the parvovirus were designed and constructed. The molecular docking and dynamic simulation analysis showed higher binding affinities and stable conformation with canine (TLR5) and feline (TLR4) toll like receptors. Though computational analysis will support us to predict the more precise top-ranked epitopes and their immuno-antigenic properties, further experimental validation will be required to be used against those viruses.

Abstract: Large language models (LLMs) and foundation models (FMs) are reshaping petroleum engineering at a pace no previous wave of artificial intelligence has matched. Between 2022 and 2026 the field went from zero petroleum specific LLMs to eighteen domain-specialized models, more than a dozen subsurface foundation models, and more than twenty commercial industry platforms, while annual publication counts grew more than five-fold from 2020 to 2024. This survey integrates those developments into a single framework. We analyze 296 verified references spanning 2003–2026 across 14 thematic areas and six petroleum sub-disciplines plus one cross-cutting category (geophysics, drilling, reservoir, production, petrophysics, completions, and cross-cutting), from classical natural-language-processing baselines through today’s vision–language models, retrieval-augmented generation stacks, and autonomous agents. Our organizing contributions include (i) a positioning matrix against 25 prior surveys, (ii) a bubble-plot taxonomy of sub-disciplines against AI paradigms, (iii) seven application-category tables, six additional thematic tables, and a dedicated maturity-model table (fourteen tables in total), (iv) a catalog of public petroleum AI systems and enabling substrate, and (v) the PetroLLM Maturity Model — a five-level scaffold (L1 Conversational Q&A, L2 Document Intelligence and Retrieval, L3 Domain-Specialized LLMs, L4 Autonomous Agents and Copilots, L5 Self-Improving Foundation-Model Ecosystems) that situates every surveyed system on a common ladder. The paper closes with a bibliometric snapshot (trends, sub-discipline distribution, method distribution, institutional footprint) and an open-research agenda spanning data, benchmarks, physics integration, safety, multilinguality, and standards. Our headline findings: geophysics leads, reservoir and production lag, petroleum benchmarks are scarce, industry deployments outpace academic publication, and L5 self-improving ecosystems remain aspirational but within a realistic 2030 horizon.

Abstract: Accurate prediction of nitrogen oxide (NOx) emissions from marine medium-speed four-stroke diesel engines is crucial for meeting increasingly stringent environmental standards. This paper focuses on optimizing the first and most significant reaction of the extended Zeldovich mechanism for the formation of nitric oxide (NO). A numerical engine model was developed and validated against experimental measurements of combustion pressure, power, and emissions at 81.95% of the Maximum Continuous Rating (MCR). The research analyzes the influence of various chemical reaction rate constants (k_1,f) on the accuracy of NO concentration predictions. The results demonstrate that by carefully selecting the kinetic parameters, the deviation of the numerical model can be reduced to only -0.93%. Utilizing the optimized constant for the primary Zeldovich reaction k_1,f = 1.8*10¹⁴ *e^(-38300/T), significantly improves the reliability of combustion and emission formation simulations.

Abstract: This study aims to present a series of cases reports of- knee arthritis occurring following COVID-19 RNA vaccination and to examine the potential role of these Spike protein-based RNA vaccines in the development of this arthritis. Although musculoskeletal disorders have been reported in connection with COVID-19 vaccines, post-vaccination arthritis is not yet listed. Given the different and specific mechanisms of Spike protein-based RNA vaccines and viral vector-based DNA vaccines, we report 7 cases that question the role of COVID-19 vaccines in the onset of early or late-onset knee arthritis observed following one or more injections. All patients (5 early onset and 2 late-onset cases) were examined in the same department and underwent a comprehensive assessment to investigate the usual causes of unilateral or bilateral oligoarthritis; none of them had a previous predisposition to rheumatic disease. No inflammatory rheumatic disease was detected, nor did any develop after a follow-up period of at least 24 months. A double-check of medical histories (both prior to and following diagnosis) was carried out by consulting the general medical database for all patients described. Post-vaccination serological monitoring of blood and synovial fluid to detect the presence of anti-spike antibodies has been requested for all patients and a synovial biopsy was performed in four of them. All patients showed improvement following low-dose prednisolone treatment within two months, but in some patients the symptoms persist. Anti-Spike antibody levels were found to be elevated in blood and synovial fluid samples from all patients. Synovial biopsies reveal chronic histiocytic (CD68+) and plasmocytic infiltration (CD3+) accompanied by neovascularisation. The similar timeline, progression and clinical manifestations of this knee arthritis might suggest a pathogenic role for the spike protein and/or an overproduction of anti-spike antibodies following mRNA vaccination.

Abstract: Cardiovascular disease (CVD), the leading cause of global mortality, is intrinsically linked to atherosclerosis. Recent research suggests that respiratory function may be associated with the development of atherosclerosis. Alterations in the composition of the respiratory microbiome can negatively affect pulmonary function, which in turn may impact cardiovascular health through mechanisms that remain incompletely understood. To analyze the upper airway microbiome abundance and diversity in adults with subclinical atherosclerosis. A case–control study was conducted for atherosclerosis. Oropharyngeal swab samples were collected from participants in the CaRes cohort, for whom carotid Doppler, spirometry, blood chemistry, and clinical history data were available. Sequencing of the 16S ribosomal RNA gene was performed, followed by comparative analyses between subjects with and without atherosclerosis. A total of 100 subjects were analyzed (50 cases and 50 controls). The phylum Bacteroidetes was the most prevalent in both groups, followed by Firmicutes and Proteobacteria. Family Enterobacteriaceae was more abundant among case group, which included species such as Serratia, Klebsiella spp., and Campylobacter. Evaluation of alpha diversity indices revealed lower levels for cases group, with a Shannon diversity index of 2.61 compared to 4.07 in controls (p = 0.006). Differences in microbiota composition were observed between cases and controls. Specifically, the family Enterobacteriaceae was associated with the presence of atherosclerosis, suggesting its potential involvement in the progression of cardiovascular disease.

Abstract:

Shinorine, a naturally occurring UV-absorbing compound belonging to the class of mycosporine-like amino acids (MAAs), has attracted considerable attention for its applications in pharmaceuticals, cosmetics, and biomaterials. However, conventional production methods based on extraction from marine organisms are constrained by low yield, limited availability, and environmental sustainability concerns. In this study, we developed a microbial cell factory for the efficient biosynthesis of shinorine in Escherichia coli. Specifically, the transaldolase gene in the pentose phosphate pathway (PPP) was precisely disrupted to block the metabolic conversion of sedoheptulose-7-phosphate (S7P), thereby enhancing its intracellular accumulation. In parallel, a cyanobacterial shinorine biosynthetic gene cluster (Ava_3858–Ava_3855) was heterologously expressed in the engineered strain, enabling the reconstruction of a functional biosynthetic pathway utilizing S7P as a key precursor. This integrated metabolic engineering strategy effectively overcomes the limitations of traditional extraction methods and significantly improves shinorine production. Moreover, the approach provides a versatile framework for the microbial synthesis of other high-value natural products, with broad implications for sustainable biomanufacturing.

Abstract: Let ξ(z) be the Riemann xi function. We prove the boundedness of coefficients of the power series expansion of ξ′(1/z)/ξ(1/z). By an observation of Keiper this implies that the Riemann hypothesis is true.

Abstract: Background/Objectives: Pregnancy is associated with anatomical, hormonal, and immunological changes that may increase susceptibility to infections and metabolic disturbances. Adequate hygiene practices and balanced nutrition, including appropriate supplementation, may help reduce maternal and fetal complications. This study aimed to describe hygiene and nutrition behaviors among women aged 18–40 years in Romania and to evaluate their association with pregnancy outcome, complications, and infections during gestation. Methods: An observational, descriptive, and analytical cross-sectional study was conducted using an anonymous online questionnaire distributed through Google Forms. Eligible participants were women aged 18–40 years from Romania who were currently pregnant or had experienced at least one previous pregnancy ending in birth or miscarriage. The questionnaire assessed socio-demographic characteristics, hand, oral, and intimate hygiene behaviors, nutritional supplementation, pregnancy-related infections and complications, and the outcome of the most recent pregnancy. Data from 61 respondents collected over a 6-month period were analyzed using descriptive statistics and Chi-square tests. Results: Most participants reported favorable hygiene practices and relatively balanced or balanced diets. Term delivery was the most frequent pregnancy outcome (75.4%), while 72.1% reported no diagnosed infection during pregnancy. A significant association was found between intimate hygiene frequency and vaginal candidiasis (χ² = 8.92; p = 0.011), with occasional intimate hygiene linked to a higher frequency of infection. Perceived diet quality was also significantly associated with pregnancy complications (χ² = 8.81; p = 0.012), and a higher level of supplementation was significantly associated with more favorable pregnancy outcomes (χ² = 28.29; p < 0.001). No significant association was found between hand-washing frequency and urinary tract infections (χ² = 2.32; p = 0.314). Conclusions: The findings suggest that adequate intimate hygiene, balanced nutrition, and appropriate nutritional supplementation during pregnancy may contribute to better maternal and fetal out-comes. These results support the importance of prenatal education and preventive strategies focused on hygiene and nutrition. Larger studies are needed to confirm these associations.

Abstract: Banana Fusarium wilt caused by Fusarium oxysporum f. sp. cubense tropical race 4 (Foc TR4) severely threatens global banana production. An endophytic Streptomyces strain 7‑1, isolated from the roots of Peliosanthes macrostegia, exhibited strong antifungal activity against Foc TR4 (inhibition rate: 79.75%) and broad-spectrum activity against 12 plant pathogenic fungi. The crude extract of strain 7‑1 inhibited Foc TR4 mycelial growth in a dose-dependent manner (EC₅₀ = 69.20 μg·mL⁻¹) by disrupting cell wall/membrane integrity, inducing mycelial damage, spore apoptosis, nucleic acid leakage and membrane lipid peroxidation. Pot experiments showed strain 7‑1 achieved 63.15% biocontrol efficacy against banana Fusarium wilt, promoted banana growth, enhanced root defense enzyme activities (POD, PPO, PAL), and regulated rhizosphere microflora by enriching beneficial microbes (Bacillus) and reducing Fusarium abundance. Metabolomic analysis identified natamycin as the major active metabolite (EC₅₀ = 8.58 μg·mL⁻¹), which exhibited similar inhibitory effects to the crude extract. Hydroponic experiments confirmed natamycin controlled banana Fusarium wilt with 34.91% efficacy at 1 × EC₅₀. In conclusion, Streptomyces sp. 7‑1 is an environmentally friendly biocontrol strain inhibiting Foc TR4 via direct pathogen damage and indirect regulation of plant defense/rhizosphere microflora. Natamycin has potential as an agricultural fungicide, providing a new candidate and theoretical basis for sustainable control of banana Fusarium wilt.

Abstract: This paper investigates whether financial markets exhibit deterministic chaotic dynamics and whether such dynamics can improve early detection of systemic crises. Using a global dataset covering major equity indices and volatility indicators over the period 1990–2025, we apply a comprehensive nonlinear framework combining maximum Lyapunov exponents, correlation dimension, sample entropy, recurrence quantification analysis (RQA), and nonlinear vector autoregressions. The results provide robust evidence that financial markets are characterized by positive Lyapunov exponents, indicating sensitive dependence on initial conditions and the presence of deterministic chaos. Importantly, chaos intensifies systematically prior to major crises, including the Dot-com bubble, the Global Financial Crisis, and the COVID-19 market crash. We construct a composite Early-Warning Index (EWI) based on nonlinear indicators, which significantly outperforms traditional benchmarks such as Value-at-Risk and volatility-based models in predicting crisis events. The findings suggest that financial instability is largely endogenous, emerging from nonlinear amplification mechanisms rather than purely exogenous shocks. By integrating chaos theory into financial econometrics, this study provides a novel framework for understanding market dynamics and offers practical tools for systemic risk monitoring. The results have important implications for macroprudential policy and the design of forward-looking early-warning systems.

Abstract: Ancient contemplative traditions offer sophisticated technologies for cultivating stable states of consciousness underlying sustainable well-being. This integrative review synthesizes empirical and theoretical literature on Patañjali’s Sanyam—the integrated practice of Dharana (concentration), Dhyana (meditative absorption), and Samadhi (non-dual awareness)—and its Zen parallel, Joriki (concentration power), with contemporary research on consciousness neuroscience, hypnosis, transpersonal psychology, and happiness science. Drawing on 187 peer-reviewed studies, this paper establishes a framework linking these practices to Fundamental Peace: a primordial, non-reactive ground state of inner stillness and clarity from which authentic happiness and flourishing emerge. Convergent evidence demonstrates that Sanyam and Joriki cultivate ekagrata (one-pointedness), producing measurable changes in brain activity, autonomic regulation, and phenomenological experience. These practices share structural parallels with clinical hypnosis and facilitate access to non-ordinary states characterized by reduced default mode network activity and dissolution of subject-object duality. Implications for transpersonal psychology, clinical practice, and global happiness initiatives are discussed.

Abstract: Background: Distinguishing genuine kinase–substrate motifs from background noise is a growing challenge, as mass spectrometry (MS) -based global phosphoproteomics identifies a rapidly expanding set of phosphorylation sites. One of the major limitations is selecting an appropriate background model that systematically controls both tech-nical and biological sources of bias. Although using the entire proteome as a back-ground in FASTA format considers the overall amino acid composition, it is still prone to biases from protein abundance and the uneven distribution of sequence space (par-ticularly around low-abundance proteins). By contrast, internal background methods can control experiment-specific detection biases, but they may not fully capture resi-due-specific compositions or general trends in phosphorylation. Methods: I develop a Dual-Background Enrichment (DBE) strategy, which involves analyzing motif en-richment against two distinct background models: (1) A residue-heterogeneous inter-nal background composed of phospho-motifs centered on the residue, e.g., serine (S) motifs, against threonine (T) and tyrosine (Y) centered motifs. (2) A FASTA back-ground that includes all canonical S, T, and Y sites from the reference proteome. Re-sults: Motifs are classified as high confidence if they meet statistical significance (q ≤ 0.05) against both background models. Conclusion: By applying the DBE strategy to a large-scale phosphoproteomics dataset, we distinguish motifs driven by amino acid composition (enriched in FASTA background only) from those reflecting kinase sub-strate specificity (enriched in both backgrounds). This dual-reference approach reduces false positives arising from sequence composition bias and identifies high-confidence candidate kinase recognition motifs.

Abstract: Medicinal plants have long served as a source of therapeutic agents in Libya and worldwide. This study investigated the in vitro antibacterial activity of twenty-two medicinal plant species against a panel of pathogenic bacteria previously isolated from food of animal origin in Libya. Plant materials were collected or purchased from different regions of Libya and extracted using three solvents of varying polarity: methanol, petroleum ether, and ethyl acetate. The extracts were screened for antibacterial activity against Escherichia coli (EC 56, EC 184), Staphylococcus aureus (SA 121), Klebsiella pneumoniae (KP 243), and Bacillus cereus (BC 4) using the agar well diffusion method. Extracts showing inhibition zones >11 mm were further evaluated for minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) using the broth microdilution method. Initial screening revealed that extracts from Thymus capitatus, Punica granatum, Syzygium aromaticum, Cinnamomum verum, Rosmarinus officinalis, and Myrtus communis exhibited notable antibacterial activity. The strongest effects were observed with petroleum ether and ethyl acetate extracts. Extracts of C. verum (R2, R3) and S. aromaticum (P2) showed the lowest MIC values, reaching 1.875 mg/mL against S. aureus (SA 121) and B. cereus (BC 4). MBC values were generally one- to two-fold higher than MIC values, indicating predominantly bactericidal effects. B. cereus (BC 4) was the most susceptible organism, whereas most extracts showed limited or no activity against E. coli strains. These findings confirm the strong antibacterial potential of several medicinal plants, particularly C. verum (cinnamon), S. aromaticum (clove), and T. capitatus (thyme). The results support their traditional use and highlight their potential as sources of natural antibacterial agents and food preservatives. Further studies are required to isolate and characterize the active compounds.

Abstract: Objective: This perspective piece examines the role of Large Language Models (LLMs) in healthcare, arguing that despite significant investment, these models have had only a limited impact. Moreover, we argue that LLMs must replicate key phases of clinical healthcare delivery to be a force multiplier, a necessary condition to address the global burden of disease. Discussion: We argue that LLMs lack the metacognitive capacity for ranked, dynamic reasoning. This is evidenced by clinically dangerous fabrications and an inability to perform unless complete information is provided. We extend clinical critiques with a statistical argument and a simulation exercise demonstrating that LLM-based diagnosis is not merely impractical but structurally incapable of converging on correct diagnoses in realistic clinical settings. Conclusion: Unless LLMs can independently collect patient history and triage, eliminate differential diagnoses, provide a treatment plan, and generate encounter notes, these models will have limited gains in efficiency relative to cognitive AI and structured reasoning approaches that are capable of functioning autonomously at each stage of the clinical workflow.

Abstract: Accurate three-dimensional representations of tree structure are essential for fire modeling, radiative transfer simulation, synthetic data generation, and digital twins of forests, yet detailed 3D structure is rarely available at required scales. Current approaches approximate crowns with smooth geometric primitives, discarding the clumping, gaps, and irregular branching present in real trees. We present TreeFlow, a conditional flow matching model that generates realistic 3D tree point clouds from species, acquisition platform, and height. The model uses a transformer trained on real laser scanning data from the FOR-species20K benchmark to learn a velocity field transporting samples from a Gaussian distribution to the source data distribution. We evaluate generation quality by comparing conditioning and distributional fidelity metrics to scans of real trees. Generated trees match or approach the intra-class baseline on five of six metrics, with a Chamfer distance of 0.581 m versus 0.559 m for real trees of the same genus and height class. Performance is strongest below 25 m and degrades with increasing height. TreeFlow is the first flow matching model to produce 3D tree point clouds from scalar inventory attributes using real laser scanning data.

Abstract: This work demonstrated improvements in the photovoltaic performance metrics of a dye-sensitized solar cell (DSSC) through the application of Eu-doped strontium silicate – Sr2SiO4:Eu3+, luminescent downshifting (LDS) material. The material converted underuti-lized high energy ultraviolet (UV) photon into lower energy visible photon for better spec-tral responsivity in the DSSC. The LDS material was prepared by the conventional solid state technique. Surface morphology was examined by scanning electron microscope (SEM). Photoluminescence (PL) measurement was applied for the fluorescence emission. The photovoltaic performances of the bare and LDS enhanced devices were analyzed from the photovoltaic current – voltage measurement. Compared to the bare DSSC, the cell with Sr2SiO4:Eu3+ LDS phosphor material had an enhancement of 14.8 % in the short circuit current density (Jsc), from 0.243 – 0.279 mA/cm2. The open circuit voltage (Voc) yielded an improvement of 10 % from 580 – 638 mV. Maximum power output (Pmax) produced a boost of 26.5 % from 0.0136 – 0.0172 mW and the efficiency improvement at 26.6 % from 1.09 – 1.38 %. The coefficient of variation was introduced to evaluate device reproducibility. The device with the incorporation of Sr2SiO4:Eu3+ LDS phosphor, depicted a coefficient of variation of 8.5 %, suggesting good DSSC reproducibility consistency.

Abstract: Computer-aided drug design (CADD) is undergoing a fundamental paradigm shift driven by the transition from classical biophysical methods to deep learning architectures and generative artificial intelligence. This review analyzes the evolution of molecular docking algorithms. We examine traditional programs (AutoDock Vina, Glide, GOLD) based on stochastic conformational search and empirical scoring functions, which retain the status of gold standard due to the high physical validity of the generated predictions. Software solutions for high-throughput virtual screening, such as distributed pipelines like EasyDock and graphical interfaces like EasyDockVina, are analyzed. Particular attention is paid to the latest generative AI models (DiffDock, GNINA, AlphaFold 3, DynamicBind, FABFlex), which address the computational challenges of blind docking and macromolecular receptor flexibility. We assess the systemic crisis of neural network generalization ability identified in independent benchmarks (PoseBusters, Bento, NextTopDocker) and substantiate the need to integrate the laws of molecular physics into the latent spaces of models. We conclude that the formation of hybrid pipelines, combining the speed of AI with the rigor of classical mechanics, is a necessary development.