Drug Discovery Research Articles

Application of machine learning based genome sequence analysis in pathogen identification

Infectious diseases caused by pathogenic microorganisms pose a serious threat to human health. Despite advances in molecular biology, genetics, computation, and medicinal chemistry, infectious diseases remain a significant public health concern. Addressing the challenges posed by pathogen outbreaks, pandemics, and antimicrobial resistance requires concerted interdisciplinary efforts. With the development of computer technology and the continuous exploration of artificial intelligence(AI)applications in the biomedical field, the automatic morphological recognition and image processing of microbial images under microscopes have advanced rapidly. The research team of Institute of Microbiology, Chinese Academy of Sciences has developed a single cell microbial identification technology combining Raman spectroscopy and artificial intelligence. Through laser Raman acquisition system and convolutional neural network analysis, the average accuracy rate of 95.64% has been achieved, and the identification can be completed in only 5 min. These technologies have shown substantial advantages in the visible morphological detection of pathogenic microorganisms, expanding anti-infective drug discovery, enhancing our understanding of infection biology, and accelerating the development of diagnostics. In this review, we discuss the application of AI-based machine learning in image analysis, genome sequencing data analysis, and natural language processing (NLP) for pathogen identification, highlighting the significant role of artificial intelligence in pathogen diagnosis. AI can improve the accuracy and efficiency of diagnosis, promote early detection and personalized treatment, and enhance public health safety.

Unraveling the pathogenic interplay between SARS-CoV-2 and polycystic ovary syndrome using bioinformatics and experimental validation

The prevalence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) among polycystic ovary syndrome (PCOS) is significantly higher than in the general population. However, the mechanisms underlying this remain obscure. This study aimed to explore the mechanisms by identifying the genetic signature of SARS-CoV-2 infection in PCOS. In the present study, a total of 27 common differentially expressed genes (DEGs) were selected for subsequent analyses. Functional analyses showed that immunity and hormone-related pathways collectively participated in the development and progression of PCOS and SARS-CoV-2 infection. Under these, 7 significant hub genes were identified, including S100A9, MMP9, TLR2, THBD, ITGB2, ICAM1, and CD86 by using the algorithm in Cytoscape. Furthermore, hub gene expression was confirmed in the validation set, PCOS clinical samples, and mouse model. Immune microenvironment analysis with the CIBERSORTx database demonstrated that the hub genes were significantly correlated with T cells, dendritic cells, mast cells, B cells, NK cells, and eosinophils and positively correlated with immune scores. Among the hub genes, S100A9, MMP9, THBD, ITGB2, CD86, and ICAM1 demonstrated potential as possible diagnostic markers for COVID-19 and PCOS. In addition, we established the interaction networks of ovary-specific genes, transcription factors, miRNAs, drugs, and chemical compounds with hub genes with NetworkAnalyst. This work uncovered the common pathogenesis and genetic signature of PCOS and SARS-CoV-2 infection, which might provide a theoretical basis and innovative ideas for further mechanistic research and drug discovery of the comorbidity of the two diseases.

Identification, characterization, and structure of a tRNA splicing enzyme RNA 5'-OH kinase from the pathogenic fungi Mucorales.

Fungal Trl1 is an essential tRNA splicing enzyme composed of C-terminal cyclic phosphodiesterase and central polynucleotide kinase end-healing domains that convert the 2',3'-cyclic-PO4 and 5'-OH ends of tRNA exons into the 3'-OH,2'-PO4 and 5'-PO4 termini required for sealing by an N-terminal ATP-dependent ligase domain. Trifunctional Trl1 enzymes are present in most human fungal pathogens and are untapped targets for anti-fungal drug discovery. Mucorales species, deemed high priority human pathogens by WHO, elaborate a noncanonical tRNA splicing apparatus in which a stand-alone monofunctional RNA ligase enzyme joins 3'-OH,2'-PO4 and 5'-PO4 termini. Here we identify a stand-alone Mucor circinelloides polynucleotide kinase (MciKIN) and affirm its biological activity in tRNA splicing by genetic complementation in yeast. Recombinant MciKIN catalyzes magnesium-dependent phosphorylation of 5'-OH RNA and DNA ends in vitro. MciKIN displays a strong preference for GTP as the phosphate donor in the kinase reaction, a trait shared with the stand-alone RNA kinase homologs from Mucorales species Rhizopus azygosporus (RazKIN) and Lichtheimia corymbifera (LcoKIN) and with the kinase domains of fungal Trl1 enzymes. We report a 1.65 Å crystal structure of RazKIN in complex with GDP•Mg2+ that illuminates the basis for guanosine nucleotide specificity.

Harnessing high potential benzothiazole chalcones against dengue virus NS5 protein: A multi-faceted theoretical study through molecular docking, ADME, and DFT

Chalcones bearing tetralone, indanone and benzothiazole cores were synthesized successfully using a general Claisen-Schmidt condensation protocol. The prepared compounds were purified and structurally analyzed by 1H, 13C NMR, and FT-IR techniques. A multi-faceted theoretical approach, combining Density Functional Theory (DFT), molecular docking, and ADME predictions, was employed to evaluate their therapeutic potential. DFT calculations at the B3LYP/def2-TZVP level revealed key electronic properties, with TD3 compound demonstrating the highest chemical reactivity. Molecular Electrostatic Potential (MEP) and Reduced Density Gradient (RDG) analyses provided insights into the compounds' non-covalent interactions and charge distributions. Molecular docking studies against the NS5 protein (PDB: 6KR2) showed superior binding affinities for all three compounds compared to the control ligand SAH, with TD3 exhibiting the lowest binding energy (−8.41 kcal/mol) and theoretical inhibition constant (689.31 nM). ADME predictions indicated favorable drug-like properties with concerns regarding aqueous solubility and potential P-glycoprotein interactions. Toxicity evaluations highlighted challenges, particularly in hepatotoxicity and carcinogenicity. The study identified TD3 as a promising lead compound for Dengue Virus NS5 inhibition, while also emphasizing the need for targeted modifications to address toxicity concerns. This research not only contributes to anti-dengue drug discovery efforts but also provides a robust methodological framework for the theoretical evaluation of similar small compounds in future investigations.

2′-Fluorinated nucleoside chemistry for new drug discovery: achievements and prospects

Abstract Fluorinated nucleosides are an important class of modified nucleosides that have demonstrated therapeutic potential for treating various human diseases, especially viral infections and cancer. Many fluorinated nucleosides have advanced into clinical trials or approved by the FDA for use in patients. Among these fluorinated nucleosides, the fluorinated nucleoside azvudine, developed by us, has been officially approved by the National Medical Products Administration (NMPA) for the treatment of coronavirus disease 2019 (COVID-19) and human immunodeficiency virus (HIV), indicating the therapeutic promise of fluorinated nucleosides. In view of the therapeutic promise of fluorinated nucleosides for antiviral and anticancer therapy, in this Review we will provide a comprehensive overview of well-established 2′-fluorinated nucleosides approved for use in the market or those in clinical stages for antiviral and antitumor therapies, highlighting the drug discovery strategies, structure-activity relationship (SAR) studies, mechanisms of actions, and preclinical/clinical studies and also discuss the challenges and future directions for nucleoside-based new drug discovery.

Advances in Drug Discovery and Design using Computer-aided Molecular Modeling.

Computer-aided molecular modeling is a rapidly emerging technology that is being used to accelerate the discovery and design of new drug therapies. It involves the use of computer algorithms and 3D structures of molecules to predict interactions between molecules and their behavior in the body. This has drastically improved the speed and accuracy of drug discovery and design. Additionally, computer-aided molecular modeling has the potential to reduce costs, increase the quality of data, and identify promising targets for drug development. Through the use of sophisticated methods, such as virtual screening, molecular docking, pharmacophore modeling, and quantitative structure-activity relationships, scientists can achieve higher levels of efficacy and safety for new drugs. Moreover, it can be used to understand the activity of known drugs and simplify the process of formulating, optimizing, and predicting the pharmacokinetics of new and existing drugs. In conclusion, computer-aided molecular modeling is an effective tool to rapidly progress drug discovery and design by predicting the interactions between molecules and anticipating the behavior of new drugs in the body.

The Metformin Immunoregulatory Actions in Tumor Suppression and Normal Tissues Protection.

The immune system is the key player in a wide range of responses in normal tissues and tumors to anticancer therapy. Inflammatory and fibrotic responses in normal tissues are the main limitations of chemotherapy, radiotherapy, and also some newer anticancer drugs such as immune checkpoint inhibitors (ICIs). Immune system responses within solid tumors including anti-tumor and tumor-promoting responses can suppress or help tumor growth. Thus, modulation of immune cells and their secretions such as cytokines, growth factors and epigenetic modulators, pro-apoptosis molecules, and some other molecules can be suggested to alleviate side effects in normal tissues and drug-resistance mechanisms in the tumor. Metformin as an anti-diabetes drug has shown intriguing properties such as anti-inflammation, anti-fibrosis, and anticancer effects. Some investigations have uncovered that metformin can ameliorate radiation/chemotherapy toxicity in normal cells and tissues through the modulation of several targets in cells and tissues. These effects of metformin may ameliorate severe inflammatory responses and fibrosis after exposure to ionizing radiation or following treatment with highly toxic chemotherapy drugs. Metformin can suppress the activity of immunosuppressive cells in the tumor through the phosphorylation of AMP-activated protein kinase (AMPK). In addition, metformin may stimulate antigen presentation and maturation of anticancer immune cells, which lead to the induction of anticancer immunity in the tumor. This review aims to explain the detailed mechanisms of normal tissue sparing and tumor suppression during cancer therapy using adjuvant metformin with an emphasis on immune system responses.

Integrating computational methods and in vitro experimental validation reveals the pharmacological mechanism of Selaginella bryopteris (L.) Baker targeting major proteins in breast cancer

Breast cancer remains a significant global health challenge, necessitating the exploration of novel therapeutic options. The present study employs an integrated approach encompassing network pharmacology, molecular docking, molecular dynamics simulations, and in-vitro validation to investigate the potential of Selaginella bryopteris in breast cancer treatment. Initial network pharmacology analysis revealed different potential targets and pathways associated with breast cancer that could be modulated by S. bryopteris phytochemical constituents. Molecular docking and dynamics simulations further elucidated the stability and dynamics of protein-ligand complexes (lanaroflavone–EGFR and sequoiaflavone-CTNNB1). The in-vitro assays demonstrated the ability of S. bryopteris crude extract to inhibit cancer cell growth (IC50 - 78.34 μg/mL) migration and invasion, supporting the computational predictions. The integrated approach employed in the present study offers a robust framework for the systematic exploration of S. bryopteris in drug discovery as a promising candidate for breast cancer treatment.

Anion-driven C[sbnd]F bond cleavage of trifluoromethyl N-aryl hydrazones toward the assembly of N-heterocycles

Herein we report an efficient and convenient method for synthesizing 1H-benzo[d]imidazoles via a base-promoted reaction of readily available α-CF3 ketone-derived hydrazones with o-phenylenediamines. The key innovation is the anion-driven triple-cleavage of the CF bond in the CF3 group, which acts as a C1 synthon at the 2-position of the 1H-benzo[d]imidazole. This novel approach simplifies the synthesis and can be extended to the preparation of benzo[d]oxazoles and benzo[d]thiazoles. Our method offers a versatile and robust platform for the synthesis of important heterocyclic compounds, with potential applications in medicinal chemistry and drug discovery

Natural History and Clinicopathological Associations of TRPC6-Associated Podocytopathy.

Understanding the genetic basis of human diseases has become integral to drug development and precision medicine. Recent advancements have enabled the identification of molecular pathways driving diseases, leading to targeted treatment strategies. The increasing investment in rare diseases by the biotech industry underscores the importance of genetic evidence in drug discovery and approval processes. Here we studied a monogenic Mendelian kidney disease, TRPC6-associated podocytopathy (TRPC6-AP), to present its natural history, genetic spectrum, and clinicopathological associations in a large cohort of patients with causal variants in TRPC6, in order to help define the specific features of disease and further facilitate drug development and clinical trials design. the study involved 64 individuals from 39 families with TRPC6 causal missense variants. Clinical data, including age of onset, laboratory results, response to treatment, kidney biopsy findings, and genetic information, were collected from multiple centers nationally and internationally. Exome or targeted sequencing was performed and variant classification was based on strict criteria. Structural and functional analyses of TRPC6 variants were conducted to understand their impact on protein function. In depth re-analysis of light and electron microscopy specimens for 9 available kidney biopsies was conducted to identify pathological features and correlates of TRPC6-AP. Large-scale sequencing data did not support causality for TRPC6 protein-truncating variants. We identified 21 unique TRPC6 missense variants, clustering in three distinct regions of the protein, and with different effects on TRPC6 3D protein structure. Kidney biopsy analysis revealed FSGS patterns of injury in most cases, along with distinctive podocyte features including diffuse foot process effacement and swollen cell bodies. The majority of patients presented in adolescence or early adulthood but with ample variation (average 22, SD ± 14 years), with frequent progression to kidney failure but with variability in time between presentation and ESKD. This study provides insights into the genetic spectrum, clinicopathological associations, and natural history of TRPC6-AP.

IS4-FAM, a fluorescent tool to study CXCR4 affinity and competitive antagonism in native cancer cells.

The ability to accurately measure drug-target interaction is critical for the discovery of new therapeutics. Classical pharmacological bioassays such as radioligand or fluorescent ligand binding assays can define the affinity or Kd of a ligand for a receptor with the lower the Kd, the stronger the binding and the higher the affinity. However, in many drug discovery laboratories today, the target of interest if often artificially upregulated by means of transfection to modify the host cell's genetic makeup. This then potentially invalidates the assumptions of classical pharmacology affinity calculations as the receptor of interest is no longer at normal physiological densities. The CXCR4 receptor is expressed on many different cancer cell types and is associated with metastasis and poor prognosis. Therefore, the CXCR4 receptor is a desirable target for novel therapeutics. In this study, we explore the applicability of the newly developed fluorescently tagged CXCR4 antagonists, IS4-FAM as an investigative tool to study CXCR4 affinity and competitive antagonism in native, non-transfected cancer cells using confocal microscopy and flow cytometry. IS4-FAM directly labels CXCR4 in several cell lines including high CXCR4 expressing SK-MEL-28 (malignant melanoma) and PC3 (metastatic prostate cancer) and lower CXCR4 expressing THP-1 (acute monocytic leukemia) and was competitive with the established CXCR4 antagonist, AMD3100. This highlights the potential of IS4-FAM as a pharmacological tool for drug discovery in native cells lines and tissues.

HTS Library Plate Rejuvenation Using a DMSO-Rich Atmosphere

Dry DMSO can rapidly pull water vapor out of the air due to its hygroscopic nature. This is a well-documented problem within drug discovery, particularly within high-throughput screening (HTS). This hydration is caused by atmospheric moisture being absorbed each time a compound library is used. This effect becomes increasingly pronounced when a compound library is used routinely. The result of this hydration is a change to both the total volume of solution and the concentration of sample still in solution. This can result in a large amount of variability in the measured biological activity of a sample depending on the library usage. In this paper, we show the detrimental effects the hydration of sample libraries has on the reproducibility of biological data and present a novel way to remove it from HTS library plates. Our approach involves creating a DMSO-rich environment, created by placing anhydrous DMSO in compound storage pods purged with nitrogen, and incubating library plates in this environment for up to 3 days. Quantification via evaporative light scattering detection (ELSD) showed that removing water greatly increased the molarity of solutions, with a greater effect being seen for compounds with poor solubility. We also demonstrated how this approach can restore the inhibitory activity of stock solutions of compounds (pIC50) of samples containing ∼30% water from >30 µM to sub-micromolar after moisture removal. This method improves the reliability of tested compounds in HTS by potentially saving pharmaceutical companies hundreds of thousands of dollars in screening campaigns and increasing the quality of data.

IPSC-derived and Patient-Derived Organoids: Applications and challenges in scalability and reproducibility as pre-clinical models

Recent advancements in stem cell technology have led to the development of organoids – three-dimensional (3D) cell cultures that closely mimic the structural and functional characteristics of human organs. These organoids represent a significant improvement over traditional two-dimensional (2D) cell cultures by preserving native tissue architecture and cellular interactions critical for physiological relevance.This review provides a comprehensive comparison between two main types of organoids: induced Pluripotent Stem Cell (iPSC)-derived and Adult Stem Cell (ASC)-derived (also known as Patient-Derived Organoids, PDOs). iPSC-derived organoids, derived from reprogrammed cells, exhibit remarkable plasticity, and can model a wide range of tissues and developmental stages. They are particularly valuable for studying early human development, genetic disorders, and complex diseases. However, challenges such as prolonged differentiation protocols and variability in maturation levels remain significant hurdles. In contrast, ASC-derived organoids, generated directly from patient tissues, faithfully recapitulate tissue-specific characteristics and disease phenotypes. This fidelity makes them indispensable for personalized medicine applications, including drug screening, disease modeling, and understanding individualized treatment responses.The review highlights the unique advantages and limitations of each organoid type, emphasizing their roles in advancing biomedical research and drug discovery. It addresses key challenges in organoid technology, such as scalability, reproducibility, and the need for standardized culture protocols. Furthermore, it explores recent innovations in scaffold-guided organoid engineering and the integration of organoids with advanced technologies like artificial intelligence and high-throughput screening.The integration of organoids with cutting-edge technologies holds promise for enhancing their utility in modeling complex human diseases and accelerating drug discovery and development. By providing more physiologically relevant models of human organs, organoid technology is poised to revolutionize biomedical research, offering new insights into disease mechanisms and personalized therapeutic strategies.

Generative AI in drug discovery and development: the next revolution of drug discovery and development would be directed by generative AI.

Generative AI in drug discovery and development: the next revolution of drug discovery and development would be directed by generative AI.

Stage III NSCLC treatment options: too many choices.

Stage III nonsmall cell lung cancer (NSCLC) represents a wide range of tumour (T1 to T4) and nodal (N0 to N3) components, requiring variable management and a multidisciplinary approach. Recent advancements in minimally invasive techniques, molecular biology and novel drug discoveries have accelerated the refinement of stage III NSCLC management. The latest developments in staging include the forthcoming update of the nodal component in the 9th TNM (tumour-node-metastasis) edition, which emphasises the critical role for endobronchial ultrasonography in mediastinal staging. Recent treatment developments include the use of immunotherapy and targeted molecular therapy in both the neoadjuvant and adjuvant setting, either in combination with other modalities or used alone as consolidation. Surgical and radiotherapy advancements have further enhanced patient outcomes. These developments have significantly improved the prognosis for patients with stage III NSCLC. Fast-changing recommendations have also brought about a challenge, with clinicians facing a number of options to choose from. Therefore, a multimodal approach by a multidisciplinary team has become even more crucial in managing stage III NSCLC.

Automation of Biochemical Assays Using an Open-sourced, Inexpensive Robotic Liquid Handler.

High Throughput Screening is crucial in pharmaceutical companies for efficient testing in drug discovery and development. Our Vaccines Analytical Research and Development Department (V-AR&D) extensively uses Robotic Liquid Handlers in their High Throughput Analytics (HTA) group for assay development and sample screening. However, these instruments are expensive and require extensive training. Opentrons' OT-2 liquid handler offers a more affordable option (< $10,000) with Python programming language and open-source flexibility, reducing training requirements. OT-2 allows broadening of testing capabilities and method transfer without significant capital investments. Two biochemical assays were conducted to assess OT-2′s performance, and it demonstrated accurate pipetting with low covariance compared to Tecan-EVO liquid handlers. Though OT-2 has some limitations such as lack of a crash detection system and limited deck space, it is a cost-effective, medium-throughput, and accurate liquid handling tool suitable for early-stage development and method transfer.

Introduction of the Trifluoropropionate Moiety into Aromatic Rings via Rhodium-Catalyzed Coupling of Arylboronic Acids with Diazoesters.

A catalytic method for the introduction of pharmaceutically meaningful fluorinated propionic acid side chains into aromatic compounds is presented. In the presence of rhodium catalyst [RhCp*Cl2]2, various arylboronic acids are efficiently coupled with an easy-to-access diazoester reagent to give perfluorinated derivatives of phenylpropionic acids, including top-selling profen drugs. The key advantage of this approach is that the pharmacophore is introduced as a whole and is compatible with various functionalities and drug discovery.

WaSPred: A reliable AI-based water solubility predictor for small molecules

A rapid and reliable evaluation of the aqueous solubility of small molecules is a hot topic for the scientific community and represents a field of particular interest in drug discovery. In fact, aqueous solubility significantly impacts various aspects that collectively influence a drug’s overall pharmacokinetics, including absorption, distribution and metabolism. For this reason, in silico approaches that provide fast and cost-effective solubility predictions, can serve as invaluable tools in the early stages of drug development. Although additional molecular features should be considered, accurate solubility predictions can help medicinal chemists rationally planning the synthesis of compounds more likely to exhibit desirable pharmacokinetic properties and in selecting the most promising candidates for further biological testing (e.g., cellular assays) from an initial pool of hit compounds with detected preliminary bioactivity. In this context, we herein report the development and evaluation of WaSPred, our AI-based water solubility predictor for small molecules. WaSPred not only showed high reliability in water solubility predictions performed on structurally heterogeneous compounds, belonging to multiple external datasets, but also demonstrated superior performance compared to a set of other commonly used water solubility predictors, thus confirming its state-of the-art robustness and its usefulness as an in silico approach for water solubility evaluations..

Rational fragment-based design of compounds targeting the PWWP domain of the HRP family

Lens epithelium-derived growth factor p75 (LEDGF/p75), member of the hepatoma-derived growth-factor-related protein (HRP) family, is a transcriptional co-activator and involved in several pathologies including HIV infection and malignancies such as MLL-rearranged leukemia. LEDGF/p75 acts by tethering proteins to the chromatin through its integrase binding domain. This chromatin interaction occurs between the PWWP domain of LEDGF/p75 and nucleosomes carrying a di- or trimethylation mark on histone H3 Lys36 (H3K36me2/3). Our aim is to rationally devise small molecule drugs capable of inhibiting such interaction. To bootstrap this development, we resorted to X-ray crystallography-based fragment screening (FBS-X). Given that the LEDGF PWWP domain crystals were not suitable for FBS-X, we employed crystals of the closely related PWWP domain of paralog HRP-2. As a result, as many as 68 diverse fragment hits were identified, providing a detailed sampling of the H3K36me2/3 pocket pharmacophore. Subsequent structure-guided fragment expansion in three directions yielded multiple compound series binding to the pocket, as verified through X-ray crystallography, nuclear magnetic resonance and differential scanning fluorimetry. Our best compounds have double-digit micromolar affinity and optimally sample the interactions available in the pocket, judging by the Kd-based ligand efficiency exceeding 0.5 kcal/mol per non-hydrogen atom. Beyond π-stacking within the aromatic cage of the pocket and hydrogen bonding, the best compounds engage in a σ-hole interaction between a halogen atom and a conserved water buried deep in the pocket. Notably, the binding pocket in LEDGF PWWP is considerably smaller compared to the related PWWP1 domains of NSD2 and NSD3 which feature an additional subpocket and for which nanomolar affinity compounds have been developed recently. The absence of this subpocket in LEDGF PWWP limits the attainable affinity. Additionally, these structural differences in the H3K36me2/3 pocket across the PWWP domain family translate into a distinct selectivity of the compounds we developed. Our top ranked compounds are interacting with both homologous LEDGF and HRP-2 PWWP domains, yet they showed no affinity for the NSD2 PWWP1 and BRPF2 PWWP domains which belong to other PWWP domain subfamilies. Nevertheless, our developed compound series provide a strong foundation for future drug discovery targeting the LEDGF PWWP domain as they can further be explored through combinatorial chemistry. Given that the affinity of H3K36me2/3 nucleosomes to LEDGF/p75 is driven by interactions within the pocket as well as with the DNA-binding residues, we suggest that future compound development should target the latter region as well. Beyond drug discovery, our compounds can be employed to devise tool compounds to investigate the mechanism of LEDGF/p75 in epigenetic regulation.

Employing Machine Learning Models to Predict Potential α-Glucosidase Inhibitory Plant Secondary Metabolites Targeting Type-2 Diabetes and Their In Vitro Validation.

The need for new antidiabetic drugs is evident, considering the ongoing global burden of type-2 diabetes mellitus despite notable progress in drug discovery from laboratory research to clinical application. This study aimed to build machine learning (ML) models to predict potential α-glucosidase inhibitors based on the data set comprising over 537 reported plant secondary metabolite (PSM) α-glucosidase inhibitors. We assessed 35 ML models by using seven different fingerprints. The Random forest with the RDKit fingerprint was the best-performing model, with an accuracy (ACC) of 83.74% and an area under the ROC curve (AUC) of 0.803. The resulting robust ML model encompasses all reported α-glucosidase inhibitory PSMs. The model was employed to predict potential α-glucosidase inhibitors from an in-house 5810 PSM database. The model identified 965 PSMs with a prediction activity ≥0.90 for α-glucosidase inhibition. Twenty-four predicted PSMs were subjected to in vitro assay, and 13 were found to inhibit α-glucosidase with IC50 ranging from 0.63 to 7 mg/mL. Among them, seven compounds recorded IC50 values less than the standard drug acarbose and were investigated further to have optimal drug-likeness and medicinal chemistry characteristics. The ML model and in vitro experiments have identified nervonic acid as a promising α-glucosidase inhibitor. This compound should be further investigated for its potential integration into the diabetes treatment system.