Findings In Models Research Articles

A molecular video-derived foundation model for scientific drug discovery

Accurate molecular representation of compounds is a fundamental challenge for prediction of drug targets and molecular properties. In this study, we present a molecular video-based foundation model, named VideoMol, pretrained on 120 million frames of 2 million unlabeled drug-like and bioactive molecules. VideoMol renders each molecule as a video with 60-frame and designs three self-supervised learning strategies on molecular videos to capture molecular representation. We show high performance of VideoMol in predicting molecular targets and properties across 43 drug discovery benchmark datasets. VideoMol achieves high accuracy in identifying antiviral molecules against common diverse disease-specific drug targets (i.e., BACE1 and EP4). Drugs screened by VideoMol show better binding affinity than molecular docking, revealing the effectiveness in understanding the three-dimensional structure of molecules. We further illustrate interpretability of VideoMol using key chemical substructures.

Quantifying tumor specificity using Bayesian probabilistic modeling for drug and immunotherapeutic target discovery.

In diseases such as cancer, the design of new therapeutic strategies requires extensive, costly, and unfortunately sometimes deadly testing to reveal life threatening off-target effects. We hypothesized that the disease specificity of targets can be systematically learned for all genes by jointly evaluating complementary molecular measurements of healthy tissues using a hierarchical Bayesian modeling approach. Our method, BayesTS, integrates protein and gene expression evidence and includes tunable parameters to moderate tissue essentiality. Applied to all protein coding genes, BayesTS outperforms alternative strategies to define therapeutic targets and nominates previously unknown targets while allowing for incorporation of new types of modalities. To expand target repertoires, we show that extension of BayesTS to splicing antigens and combinatorial target pairs results in more specific targets for therapy. We expect that BayesTS will facilitate improved target prioritization for oncology drug development, ultimately leading to the discovery of more effective and safer treatments.

Practical approach for sand-shale mixtures classification based on rocks multi-physical properties

Sandstones are the most common reservoir rocks, providing reservoirs for oil and gas and serving as reservoirs for groundwater. The Gulf of Mexico is known for its sand-shale mixtures and potential for its oil and hydrate gas resources in sandstone units. Understanding these variations is essential for assessing hydrocarbon potential and unconventional prospectivity. In this study, we utilized the Elastic, Electrical, and Radioactive (EER) properties of rocks for lithological categorization of well logging data, leading to the development of a novel rock physics template. The electrical and radioactive properties of the rocks facilitated a broad lithological classification, while their elastic characteristics helped distinguish between porous and low-porosity zones. Electrical and radioactive properties are utilized for well data classification because in sandstone formations, there is a decrease in log gamma and an increase in log resistivity. As a result, these opposing shifts in the two geophysical logs enhance the spread of data points on the lithological resistivity-gamma ray scatter plot, thereby simplifying the process of lithological categorization. Ultimately, the well logging data was sorted into three distinct categories: low shale sands (shale volume < 30 %), sand-shale mixtures (shale volume = 30 to 80 %), and shale-dominated areas. Subsequently, the Thomas Stieber model was employed to identify the types of clay minerals present in both sandstones and sand-shale mixtures. The model's findings revealed that dispersed type clay minerals are predominantly found in sandstones, with laminar and structured types being relatively rare. However, in sand-shale mixtures, both dispersed and laminar clays observed.

Transitive reasoning: A high-performance computing model for significant pattern discovery in cognitive IoT sensor network

Current research on the Internet of Things (IoT) has given rise to a new field of study called cognitive IoT (CIoT), which aims to incorporate cognition into the designs of IoT systems. Consequently, the CIoT inherits specific attributes and challenges from IoT. The CIoT applications generate vast, diverse, constantly changing, and time-dependent data due to the billions of devices involved. The efficient operation of these CIoT systems requires the extraction of valuable insights from vast data sources in a computationally efficient manner. Therefore, this study proposes transitive reasoning to glean significant concepts and patterns from a 21.25-year environmental dataset. To reduce the effects of missing entries, the proposed methodology includes a grouping of data using probabilistic clustering and applying total variance regularization in the alternate direction method of multipliers (ADMM) to regularize the sensory data. As a result, noisy entries will be less conspicuous. Afterward, it calculates the transitional plausibility value for each cluster using the transited value and then turns it into binary data to create concept lattices. In addition, each concept that is formed is assigned a weight, and the concept with the largest transitive strength value is chosen, followed by calculating the mean value. Therefore, this pattern is seen as significant. Experimental results on 21.25-year environmental data show an accuracy of over 99.5%, outperforming competing methods, as shown by cross-validation using multiple metrics.

Predicting transcriptional responses to novel chemical perturbations using deep generative model for drug discovery

Understanding transcriptional responses to chemical perturbations is central to drug discovery, but exhaustive experimental screening of disease-compound combinations is unfeasible. To overcome this limitation, here we introduce PRnet, a perturbation-conditioned deep generative model that predicts transcriptional responses to novel chemical perturbations that have never experimentally perturbed at bulk and single-cell levels. Evaluations indicate that PRnet outperforms alternative methods in predicting responses across novel compounds, pathways, and cell lines. PRnet enables gene-level response interpretation and in-silico drug screening for diseases based on gene signatures. PRnet further identifies and experimentally validates novel compound candidates against small cell lung cancer and colorectal cancer. Lastly, PRnet generates a large-scale integration atlas of perturbation profiles, covering 88 cell lines, 52 tissues, and various compound libraries. PRnet provides a robust and scalable candidate recommendation workflow and successfully recommends drug candidates for 233 diseases. Overall, PRnet is an effective and valuable tool for gene-based therapeutics screening.

A microvascularized in vitro liver model for disease modeling and drug discovery

Drug discovery for complex liver diseases faces alarming attrition rates. The lack of non-clinical models that recapitulate key aspects of liver (patho)-physiology is likely contributing to the inefficiency of developing effective treatments. Of particular notice is the common omission of an organized microvascular component despite its importance in maintaining liver function and its involvement in the development of several pathologies. Increasing the complexity ofin vitromodels is usually associated with a lack of scalability and robustness which hinders their implementation in drug development pipelines. Here, we describe a comprehensive liver microphysiological system comprising stellates, liver-derived endothelial cells and hepatocytes conceived within a scalable and automated platform. We show that endothelial cells self-organize in a microvascular network when co-cultured with stellates in a hydrogel. In a tri-culture, hepatocytes polarize accordingly, with a basolateral side facing blood vessels and an apical side facing bile-canaliculi-like structures. Stellates interact and surround the hollow microvessels. Steatosis was induced by exogenous administration of fatty acids which could be prevented by co-administration of firsocostat. Administration of TGF-βresulted in an activated stellate cells phenotype which could be prevented by the co-administration of SB-431542. The model was implemented on a microtiter plate format comprising 64 chips which enabled the development of a fully automated, multiplexed fibrosis assay with a robust Z' factor suitable for high-throughput applications.

Zebrafish: A Versatile Animal Model in Biomedical Research

Zebrafish (Danio rerio) has emerged as a powerful animal model in biomedical research, offering a unique combination of genetic tractability, rapid development, and conserved gene function with humans. This review highlights the significance of zebrafish in understanding various human diseases, including neurodegenerative disorders, metabolic disorders, cancer, and more. The zebrafish genome, with over 70% similarity to the human genome, allows for the modelling of human diseases with high fidelity. The ease of genetic manipulation, high fecundity, and low maintenance costs make zebrafish an attractive model for large-scale genetic screens and drug discovery. Zebrafish models of neurodegenerative diseases, such as Parkinson's disease, Alzheimer's disease, and Huntington's disease, have been developed, providing insights into disease mechanisms and potential therapeutic targets. Similarly, zebrafish models of metabolic disorders, including diabetes, dyslipidaemia, and atherosclerosis, have been established, enabling the study of disease pathogenesis and prevention strategies.The zebrafish has also been used to model various types of cancer, including leukaemia, melanoma, and pancreatic cancer, facilitating the discovery of novel oncogenes and tumor suppressors. The transparent nature of zebrafish embryos and larvae allows for in vivo imaging of cancer cells, enabling the study of cancer cell behavior and response to therapy. Furthermore, zebrafish has been used to investigate brain-to-organ communication, cell transplantation, and cell-cell interactions, providing insights into the complex interactions between different tissues and organs. The development of novel tools, such as CRISPR/Cas9 genome editing and single-cell RNA sequencing, has further expanded the capabilities of zebrafish research.

Exploring Metarhabditis blumi as a Model for Anthelmintic Drug Discovery

Exploring Metarhabditis blumi as a Model for Anthelmintic Drug Discovery

Perturbation Theory Machine Learning Model for Phenotypic Early Antineoplastic Drug Discovery: Design of Virtual Anti-Lung-Cancer Agents

Perturbation Theory Machine Learning Model for Phenotypic Early Antineoplastic Drug Discovery: Design of Virtual Anti-Lung-Cancer Agents

Data-augmented machine learning scoring functions for virtual screening of YTHDF1 m6A reader protein

Machine learning is rapidly advancing the drug discovery process, significantly enhancing speed and efficiency. Innovation in computer-aided drug design is primarily driven by structure- and ligand-based approaches. When the number of known inhibitors for a target is limited, data augmentation strategies are often preferred to enhance model performance. In this study, we developed predictive machine learning models for structure-based drug discovery leveraging multiple traditional machine learning algorithms trained with target and ligand dynamics-aware datasets.To illustrate our approach, we present a composite model that combines classification and regression to predict YTHDF1 inhibitors, utilizing PLEC features. YTHDF1, a key m6A reader protein involved in mRNA translation, is implicated in various cancers, making it a promising therapeutic target. Traditional structure-based virtual screening (SBVS) using generic scoring functions has struggled to identify potent YTHDF1 inhibitors due to the protein's unique binding characteristics. To overcome this, we developed YTHDF1-specific machine learning scoring functions (MLSFs) to enhance SBVS efficacy.We employed various data augmentation techniques to generate a comprehensive dataset, incorporating multiple conformations of ligands and the YTHDF1 protein. We have trained 64 YTHDF1-specific MLSFs using four machine learning algorithms and evaluated them on ten test sets, focusing on their predictive and ranking power. Our results demonstrate that the artificial neural network with protein-ligand extended connectivity fingerprints (ANN-PLEC) outperforms other MLSFs, consistently achieving high area under the precision-recall curve (PR-AUC) of 0.87. This method shows promise for targets with limited quantities of active molecules, providing a viable path forward for drug discovery research. The ANN-PLEC scoring function is made freely available on GitHub for other researchers to access and utilize https://github.com/JuniML/SBVS-YTHDF1/.

Artificial intelligence alphafold model for molecular biology and drug discovery: a machine-learning-driven informatics investigation.

AlphaFold model has reshaped biological research. However, vast unstructured data in the entire AlphaFold field requires further analysis to fully understand the current research landscape and guide future exploration. Thus, this scientometric analysis aimed to identify critical research clusters, track emerging trends, and highlight underexplored areas in this field by utilizing machine-learning-driven informatics methods. Quantitative statistical analysis reveals that the AlphaFold field is enjoying an astonishing development trend (Annual Growth Rate = 180.13%) and global collaboration (International Co-authorship = 33.33%). Unsupervised clustering algorithm, time series tracking, and global impact assessment point out that Cluster 3 (Artificial Intelligence-Powered Advancements in AlphaFold for Structural Biology) has the greatest influence (Average Citation = 48.36 ± 184.98). Additionally, regression curve and hotspot burst analysis highlight "structure prediction" (s = 12.40, R2 = 0.9480, p = 0.0051), "artificial intelligence" (s = 5.00, R2 = 0.8096, p = 0.0375), "drug discovery" (s = 1.90, R2 = 0.7987, p = 0.0409), and "molecular dynamics" (s = 2.40, R2 = 0.8000, p = 0.0405) as core hotspots driving the research frontier. More importantly, the Walktrap algorithm further reveals that "structure prediction, artificial intelligence, molecular dynamics" (Relevance Percentage[RP] = 100%, Development Percentage[DP] = 25.0%), "sars-cov-2, covid-19, vaccine design" (RP = 97.8%, DP = 37.5%), and "homology modeling, virtual screening, membrane protein" (RP = 89.9%, DP = 26.1%) are closely intertwined with the AlphaFold model but remain underexplored, which implies a broad exploration space. In conclusion, through the machine-learning-driven informatics methods, this scientometric analysis offers an objective and comprehensive overview of global AlphaFold research, identifying critical research clusters and hotspots while prospectively pointing out underexplored critical areas.

Drug Repurposing Using Computational Tools and Preventive Strategies for COVID-19 and Future Pandemics

Coronavirus disease-19 (COVID-19) was declared a global pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that targets the lower respiratory tract in human species. Since the target receptor molecule is angiotensin-converting enzyme 2 (ACE2), which is highly expressed on the cell lining of the respiratory tract, the virus is responsible for causing acute respiratory distress syndrome (ARDS). Computational tools such as artificial intelligence, machine learning and deep learning-based platforms are used for drug repurposing, identification, and selection of target molecules that can be used for vaccine development and antiviral drug production. Computer-aided drug designing, molecular docking and homology modelling are some of the most widely used in silico models for drug discovery against COVID-19. It is important to take preventive measures to control the spread of the virus. The present review focuses on the computational tools used for recognising the target molecules for vaccine production, the transmission networks of COVID-19 and the long-term strategies to prevent future pandemics such as COVID-19.

6577 National Trends in Fragility Hip Fractures Among Community Dwelling Women With Low Femoral Neck Bone Mineral Density

Abstract Disclosure: A. Atri: None. C. Anastasopoulou: None. Introduction: Fragility hip fractures are a major public health burden and are considered a good measure of the quality of healthcare. However, contemporary trends in fragility hip fractures in women with low femoral neck bone mineral density (BMD) in the United States are not known. Methods: We used data from the National Health and Nutrition Examination Survey (NHANES) for the years 2009-10, 2013–14 and 2017–18 to include U.S. women aged ≥ 60 years with bone mineral density (BMD) T score ≤ -1 at the femur neck, which was measured by dual energy x-ray absorptiometry (DEXA) on Hologic Discovery Model A Densitometers (Hologic, Inc., Bedford, Massachusetts). Fragility fracture at the hip was defined as self-reported hip fracture resulting from a fall from standing height or less. Participants with missing BMD and fracture data were excluded. ANOVA was used for comparing means. The Cochran-Armitage test was utilized for trend analysis. Results: Overall, in women aged ≥ 60 years with femoral neck T score ≤ -1, the average age of women of low BMD decreased from 71(±7.2) years to 69.6 (±6.7) years, whereas the mean T-score modestly worsened from – 1.96 (±0.62) to -2.04 (±0.77) from 2009-10 to 2017-18. Similarly, the prevalence of women with T score ≤ -2.5 increased from 18.1% to 21.3%. We also found no significant change in proportion of women who reported treatment for osteoporosis (2009-10: 19.8%, 2013-14: 18.5%, 2017-18: 19.7%; p trend = 0.052). However, proportion of women who experienced fragility hip fractures decreased significantly (2009-10: 2.4%, 2013-14: 1.5%, 2017-18: 1.2%; p trend = 0.001). The mean age of index fragility hip fracture was 70.6 years in 2009-10, 68.3 years in 2013-14 and 71.7 years in 2017-18 (p = 0.001). Age stratified analysis showed a decreasing trend in proportion of women ≥ 80 years who experienced their index fragility hip fracture (2009-10: 25.9%, 2013-14: 14.7%, 2017-18: 15%; p trend = 0.001) and 70-79 years (2009-10: 40.9%, 2013-14: 21.1%, 2017-18: 35.8%; p trend = 0.001), while an increasing trend was seen in women 60–69 years (2009-10: 33.2%, 2013-14: 64.2%, 2017-18: 49.2%; p trend = 0.001). Discussion In this nationally representative cohort of community-dwelling US women with a low femoral neck BMD, we found a 50% decrease in burden of fragility hip fractures over a 10-year period from 2009-10 to 2017-18. This trend was observed despite worsening mean femoral neck T-scores, T score ≤ -2.5, and similar rates of osteoporosis treatment. Interestingly, the proportion of women who experienced their index fragility hip fracture showed a declining trend among the 70-79 years and ≥ 80 years age-groups. This change may be due to increasing awareness and utilization of measures to decrease falls such as exercise, nutrition, health education and environment modifications targeted towards the elderly population. Presentation: 6/2/2024

9288 National Trends in Fragility Hip Fractures Among Community Dwelling Women With Low Femoral Neck Bone Mineral Density

Abstract Disclosure: A. Atri: None. C. Anastasopoulou: None. Introduction: Fragility hip fractures are a major public health burden and are considered a good measure of the quality of healthcare. However, contemporary trends in fragility hip fractures in women with low femoral neck bone mineral density (BMD) in the United States are not known. Methods: We used data from the National Health and Nutrition Examination Survey (NHANES) for the years 2009-10, 2013–14 and 2017–18 to include U.S. women aged ≥ 60 years with bone mineral density (BMD) T score ≤ -1 at the femur neck, which was measured by dual energy x-ray absorptiometry (DEXA) on Hologic Discovery Model A Densitometers (Hologic, Inc., Bedford, Massachusetts). Fragility fracture at the hip was defined as self-reported hip fracture resulting from a fall from standing height or less. Participants with missing BMD and fracture data were excluded. ANOVA was used for comparing means. The Cochran-Armitage test was utilized for trend analysis. Results: Overall, in women aged ≥ 60 years with femoral neck T score ≤ -1, the average age of women of low BMD decreased from 71(±7.2) years to 69.6 (±6.7) years, whereas the mean T-score modestly worsened from – 1.96 (±0.62) to -2.04 (±0.77) from 2009-10 to 2017-18. Similarly, the prevalence of women with T score ≤ -2.5 increased from 18.1% to 21.3%. We also found no significant change in proportion of women who reported treatment for osteoporosis (2009-10: 19.8%, 2013-14: 18.5%, 2017-18: 19.7%; p trend = 0.052). However, proportion of women who experienced fragility hip fractures decreased significantly (2009-10: 2.4%, 2013-14: 1.5%, 2017-18: 1.2%; p trend = 0.001). The mean age of index fragility hip fracture was 70.6 years in 2009-10, 68.3 years in 2013-14 and 71.7 years in 2017-18 (p = 0.001). Age stratified analysis showed a decreasing trend in proportion of women ≥ 80 years who experienced their index fragility hip fracture (2009-10: 25.9%, 2013-14: 14.7%, 2017-18: 15%; p trend = 0.001) and 70-79 years (2009-10: 40.9%, 2013-14: 21.1%, 2017-18: 35.8%; p trend = 0.001), while an increasing trend was seen in women 60–69 years (2009-10: 33.2%, 2013-14: 64.2%, 2017-18: 49.2%; p trend = 0.001). Discussion In this nationally representative cohort of community-dwelling US women with a low femoral neck BMD, we found a 50% decrease in burden of fragility hip fractures over a 10-year period from 2009-10 to 2017-18. This trend was observed despite worsening mean femoral neck T-scores, T score ≤ -2.5, and similar rates of osteoporosis treatment. Interestingly, the proportion of women who experienced their index fragility hip fracture showed a declining trend among the 70-79 years and ≥ 80 years age-groups. This change may be due to increasing awareness and utilization of measures to decrease falls such as exercise, nutrition, health education and environment modifications targeted towards the elderly population. Presentation: 6/2/2024

6578 National Trends in Vitamin D Deficiency and Racial-Ethnic Differences Among US Women With Low Bone Mineral Density

Abstract Disclosure: A. Atri: None. C. Anastasopoulou: None. Introduction: Vitamin D (Vit D) plays a critical role in maintaining bone mineralization and its deficiency adversely impacts bone health. Contemporary trends of Vit D deficiency (VDD) and racial-ethnic differences in these trends among US women with low bone mineral density (BMD) are not well elucidated. Methods: We used data from the National Health and Nutrition Examination Survey (NHANES) for the years 2009-10, 2013–14 and 2017–18 to include U.S. women ≥ 50 years with femoral neck T score ≤ -1. Osteoporosis was defined by a T-score ≤ -2.5. BMD assessment in NHANES was not completed in 2011–12 and 2015–16, and therefore not included in the analysis. Femoral neck BMD was measured by dual energy x-ray absorptiometry (DEXA) on Hologic Discovery Model A Densitometers (Hologic, Inc., Bedford, Massachusetts). VDD was defined as serum 25(OH)D &amp;lt;30 nmol/L. Participants with missing data on race-ethnicity, BMD and Vit D data were excluded. Independent samples median test was used for comparing medians. The Cochran-Armitage test was utilized for trend analysis. Results: Overall, in women with low BMD, from 2009-10 to 2017-18, prevalence of osteoporosis increased from 13.7% to 16.9% while VDD decreased significantly from 5.6% to 2.3% (p trend = 0.001). When stratified by race-ethnicity, the prevalence of osteoporosis increased in Whites (14.1% to 17.6%, p trend = 0.001), Hispanics (12.6% to 13.7%, p trend = 0.001) and Other race-ethnicity (Asian/Mixed) (14.1% to 21.4%, p trend = 0.001), but a modest decrease was noted among the Black population (7.3% to 6.8%, p trend = 0.001). VDD was disproportionately high among Blacks (In 2009-10: 28.1% vs 4.2% in Whites, 5.4% in Hispanics and 10.7% among Others). However, significantly improving trends in VDD were observed among all the groups, regardless of race/ethnicity, to 12.5% prevalence among Blacks, 1.3% in Whites, 4.9% in Hispanics and 1.1% in the Other race ethnicity group (p trend = 0.001). Discussion In this contemporary analysis, rates of VDD have decreased from nearly 1 in 20 to 1 in 50 women with low BMD in the US. Highest prevalence of VDD was noted among Blacks, seen in nearly 1 in 3 Black women in 2009-10, with improvement to 1 in 8, in 2017-18. Similar favorable trends of decline in VDD were observed across all other racial-ethnic groups as well. Despite the overall improvement in VDD, prevalence of osteoporosis declined only among Black women, suggesting the potential role of other protective factors against osteoporosis among them. Nevertheless, VDD continues to remain prevalent in this high-risk group, emphasizing the need for public health interventions to bridge this gap. Presentation: 6/2/2024

Drug Discovery and Homology Modeling Study of AKR4C14, an Aldo- Keto Reductase from Thai Jasmine Rice

Drug Discovery and Homology Modeling Study of AKR4C14, an Aldo- Keto Reductase from Thai Jasmine Rice

666P MyoScreen™ infantile and late-onset Pompe human skeletal muscle disease models for drug discovery and development of next-generation gene therapies

666P MyoScreen™ infantile and late-onset Pompe human skeletal muscle disease models for drug discovery and development of next-generation gene therapies

287 (PB275): Developing novel AR mutant prostate cancer models for CRPC related drug discovery

287 (PB275): Developing novel AR mutant prostate cancer models for CRPC related drug discovery

Utilization of an optimized AlphaFold protein model for structure-based design of a selective HDAC11 inhibitor with anti-neuroblastoma activity.

AlphaFold is an artificial intelligence approach for predicting the three-dimensional (3D) structures of proteins with atomic accuracy. One challenge that limits the use of AlphaFold models for drug discovery is the correct prediction of folding in the absence of ligands and cofactors, which compromises their direct use. We have previously described the optimization and use of the histone deacetylase 11 (HDAC11) AlphaFold model for the docking of selective inhibitors such as FT895 and SIS17. Based on the predicted binding mode of FT895 in the optimized HDAC11 AlphaFold model, a new scaffold for HDAC11 inhibitors was designed, and the resulting compounds were tested in vitro against various HDAC isoforms. Compound 5a proved to be the most active compound with an IC50 of 365 nM and was able to selectively inhibit HDAC11. Furthermore, docking of 5a showed a binding mode comparable to FT895 but could not adopt any reasonable poses in other HDAC isoforms. We further supported the docking results with molecular dynamics simulations that confirmed the predicted binding mode. 5a also showed promising activity with an EC50 of 3.6 µM on neuroblastoma cells.

A novel discovery model for revealing substitution relationships from international stock markets: With association rule analysis

At present, the trading volume of the stock market is huge, and the traditional method can not effectively find the relationship between the rise and fall of the stock market, but the machine learning method can find their interrelated data from a large number of data. This research aims to determine the effectiveness of association mining technology in analyzing the relationship between the ups and downs of stock markets in various countries, and it found the highest level of association between stock market items as investor references. The research data takes Taiwan's stock market as the target market and the international mainstream stock index as the related stock market. Through the analysis, it is found that association mining can accurately find the associated stock market according to the relevant parameters. The Taiwan stock market is more closely related to the top ten economies such as the Mainland, the United States, the United Kingdom and France, which shows that the rise of the international or mainland stock market will drive foreign capital to actively buy the Taiwan stock market, and vice versa. At last, the study sorted out three groups of stocks with the highest correlation degree according to the results of association mining, Namely Foxconn Stock (2354) and TSMC (2330), which are most closely related to the rise and fall of the international stock market. Therefore, the results of this study can also be used as a reference for investors to choose the stock price of Taiwan stock market.