Introduction

The skeletal system plays a fundamental role in maintaining overall human health, supporting movement, protecting vital organs, and facilitating physiological functions such as blood cell production and mineral storage. Throughout the life cycle, the skeletal system undergoes dynamic changes—from rapid growth in childhood1 to stabilization in adulthood and potential degeneration with aging2. As individuals age, bone mineral density (BMD) decreases, and conditions like osteoporosis become more prevalent3,4, which urgently requires strengthening research on factors related to orthopedic health. Bone health can be influenced by various factors5,6,7,8, including nutrient intake9, and numerous studies have reported associations between specific nutrients and orthopedic diseases10,11. For instance, insufficient calcium intake is strongly associated with an increased risk of osteoporosis12, while a study has shown a significant association between serum trace elements and BMD, noting that excessive iron levels may lead to bone loss13. Proper vitamin supplementation is also crucial for maintaining bone health. A review study points out that vitamins C, E, and K2 promote osteoblast development via pathways such as BMP/SMAD, TGF-β/SMAD, and Wnt/β-catenin14. Additionally, vitamin K2 increases osteoprotegerin levels15 and prevents vascular calcification16. Maintaining bone health also requires a reasonable intake of dietary protein; for example, an article in The Lancet Diabetes & Endocrinology suggests a positive correlation between BMD and total protein intake17.

Food serves as the primary source of these critical nutrients, providing essential macronutrients (e.g., proteins, carbohydrates, fats) and micronutrients (e.g., vitamins, minerals) that support bodily functions. A balanced and diverse diet not only enriches daily life but also lays the foundation for optimal health, aligning with the concept of “food as medicine” advocated by the scientific community18. Adjusting dietary patterns has been proposed as a preventive strategy to enhance physical fitness and reduce disease risk19. Given the established links between diet and bone health, there is a pressing need to explore the causal relationships between specific dietary components and orthopedic conditions. Current research, remains limited in its ability to establish clear causal connections due to confounding factors in observational studies. To address this limitation, Mendelian randomization (MR) analysis provides an effective approach to investigating how dietary factors influence skeletal health via utilizing genetic variants as instrumental variables. In a study, Feng et al. investigated the 24-hour coffee consumption of 3,676 osteoarthritis patients through a structured questionnaire and conducted MR analyses using multiple methods, including inverse variance weighting (IVW). Their findings revealed a significant positive correlation between coffee consumption and an elevated risk of osteoarthritis20. Another MR study focusing on chronic pain (including knee pain) showed that the intake of cheese, cereals, dried fruits, and fresh fruits was negatively correlated with pain, whereas the consumption of pork and poultry was positively correlated with pain21. However, current research tends to focus narrowly on the relationship between specific types of dietary intake and single orthopedic diseases, leading to a relatively limited and non-systematic analytical perspective. Additionally, the presentation of data in existing studies is predominantly confined to traditional tabular formats, lacking intuitive visualizations of the raw computational results. This not only prevents the full value of the original data from being realized but also restricts the dissemination, sharing, and secondary development and application of the data. These issues hinder a comprehensive understanding of the complex relationships between diet and orthopedic health.

To systematically explore the potential associations between dietary intake and orthopedic disorders, we developed a MR analysis database named ONMR (https://onmr.ai-bio.net). This database integrates 210 common types of dietary intake and 503 bone-related terminology descriptions, making it the most comprehensive resource currently available for MR analyses linking dietary intake to orthopedic disorders. Through the ONMR database, users can conveniently query the potential risk impacts of a specific food on various orthopedic diseases and can also perform one-click searches to identify which food intakes may influence a specific orthopedic disease. Based on the analysis using ONMR, we uncovered the impact duality of dietary consumption and the age-dependent relationship between diet and bone density. Furthermore, all query results support direct download of raw data, facilitating personalized analyses for researchers. The ONMR database presents data in a comprehensive and unbiased manner, providing valuable resources for interdisciplinary research at the intersection of food science and orthopedic science. The development of ONMR not only aids in deepening the understanding of the complex relationships between diet and skeletal health but also promotes the advancement of related fields toward more precise and personalized directions.

Results

ONMR database constructed based on large-scale Mendelian randomization analysis

This study utilized 210 dietary intake records (nutrition items) from UK Biobank as exposure factors and 503 orthopedic disorders terms obtained from UK Biobank, FinnGen, GWAS Catalog, and other GWAS datasets as outcome factors (Fig. 1A). Through the implementation of over one hundred thousand MR calculations (Fig. 1B), a dataset encompassing more than 2.7 million pairs of beta effect values was generated, including single SNP analysis results and comprehensive analyses using various methods. Aforementioned data were used to construct the ONMR database, a bioinformatics platform that which systematically maps causal relationships for different dietary intake scenarios in relation to orthopedic disorders (Fig. 1C). The ONMR database homepage supported fuzzy matching searches, allowing users to easily query specific food-related risk diseases or identify risk food intakes associated with specific diseases. The ONMR website offered dynamic selection features, enabling users to customize and draw forest plots online to showcase potential causal relationships between dietary intake and orthopedic disorders. All charts were based on beta effect values and came with options for downloading raw data (in. json format) and. svg vector images, catering to personalized user needs. Built on HTTPS architecture, ONMR operated without user registration or login, data uploads, or tracking cookies, ensuring secure access to nutrient-bone interaction networks with privacy.

Fig. 1: Construction Process of the ONMR Database.
figure 1

A Data Retrieval and Grouping Strategy: GWAS data were primarily sourced from the UK Biobank, FinnGen, and EBI. The filtering stage focused on GWAS-ID de-duplication and additional text processing to ensure data integrity. B Mendelian Randomization Calculation Steps: The TwoSampleMR package was employed for two-sample Mendelian randomization analysis, conducting a total of 105,630 (210 × 503) calculations to identify all potential associations between food intake and orthopedic diseases. During the exposure factor calculation, the primary parameter used was p1 = 5e-6, which ensured the selection of SNPs significantly correlated with exposure factors as instrumental variables (IVs). For outcome factor analysis, high correlation between SNPs and their proxy SNPs was maintained by setting proxies=TRUE and rsq=0.8. In the MR analysis, multiple methods—Inverse Variance Weighted, MR Egger, Weighted Median, Weighted Mode, and Simple Mode—were utilized to estimate causal effects, enhancing the robustness of the results. C Schematic Diagram of the ONMR Database Homepage: This diagram provides an overview of the database’s main interface, illustrating how users can navigate and interact with the comprehensive dataset.

Full size image

The impact duality of dietary intake on orthopedic health

After analyzing all Inverse variance weighted (IVW) data, we retained 4,858 MR results that reached statistical significance (p < 0.05), with 100 of these showing high statistical significance (p < 0.0001). These results covered 18 types of dietary intake. To explored the associations between the foods and orthopedic disorders, we conducted a comprehensive literature review. We found that although most foods lacked direct studies, common foods such as Beef, Bread, Cereal, Cheese, Coffee, Dried fruit, and Oily fish had been the subject of some clinical observations, animal experiments, or other Mendelian analyses (Table 1). Generally, when the beta coefficient (β) > 0, it indicates a positive correlation between the exposure and the outcome. Conversely, if β < 0, it suggests a negative correlation between the exposure factor and the outcome factor. As shown in Table 1, the computational results from ONMR indicated that the intake of beef, coffee, and poultry was positively correlated with Gonarthrosis, Osteoarthritis, and Knee pain, respectively. This meant that as the consumption of these foods increased, the risk of developing the corresponding diseases also rose. Correspondingly, the intake of other foods listed in the Table 1 (such as bread, cereal, and cheese) was negatively correlated with the risk of orthopedic diseases. The calculated results of dietary intake and disease risk listed in Table 1 were fully consistent with the trends identified by previous studies. This indicated that the conclusions drawn from the ONMR analysis had high reliability and external validity.

Table 1 Significant MR analysis results of food intake and orthopedic diseases
Full size table

Additionally, in a comprehensive assessment of the MR data, we found that the top 10 items with the smallest p-values were all related to cheese intake, with cheese intake significantly reducing the risk of Knee pain, Neck or shoulder pain, osteoarthritis, and Gonarthrosis (Supplementary Table 1). Meanwhile, the top 10 items with the highest absolute values of β revealed different trends: intake of Vegetarian alternatives increased the risk of Juvenile rheumatoid arthritis but decreased the risk of other disorders of synovium, tendon, and bursa; whereas intake of Lobster/crab, while lowering the risk of Drug-induced osteoporosis, might increase the risk of Bone and articular cartilage issues (Table 2).

Table 2 Top 10 food intakes ranked by absolute effect sizes (β) and their associations with orthopedic diseases
Full size table

The MR analysis results indicated that dietary intake had a significant dual effect on orthopedic disorders, with each food showing high-risk effects in at least one disease category while exhibiting protective effects in another. Further analysis revealed that six foods were particularly notable in this regard, as they could substantially increase the risk of certain diseases (β > 10) while significantly reducing the incidence of others (β < -10). The six foods were Lobster/crab, Naan bread, Other cheese, Other meat, Shellfish, and Vegetarian alternatives. Browsing the IVW (Nutri) webpage in our database, we generated forest plots illustrating the relationships between these foods and orthopedic disease risks (Fig. 2), to visually demonstrated the complex patterns of food intake across different diseases. As shown in Fig. 2, Lobster/crab intake reduced the occurrence of drug-induced osteoporosis but increased the risk of osteochondrodysplasias; Shellfish intake lowered the risk of ankylosis of joint, but increased the risk of other intervertebral disk disorders; and the consumption of other meats decreased alkaline phosphatase levels but correlated positively with injuries of neck.

Fig. 2: Typical foods with dual impact on orthopedic diseases.
figure 2

The figure sequentially illustrates the impact of “Lobster/crab,” “Naan bread,” “Other cheese,” “Other meat,” “Shellfish,” and “Vegetarian alternatives” on orthopedic diseases (Disorders). Each type of food intake is associated with an increase in risk for at least one disease and a decrease in risk for another. Red lines in the figure indicate an increased risk of disease following the consumption of these foods, while blue lines denote a decreased risk. All displayed data are typical inverse variance weighted (IVW) calculation results, with effect sizes (|β |) greater than 10, highlighting the significant dual impact of these foods on orthopedic diseases.

Full size image

Age-dependent relationship between diet and bone mineral density: from promotion to suppression

To figure out the mechanisms by which nutrition items affected orthopedic health, this study also investigated the differential impacts of various food intakes on the same disease—using BMD as an example—across different age groups. Based on GWAS data, we divided the samples into six subgroups by age: 0–15 years, 15–30 years, 30–45 years, 45–60 years, over 60 years, and all ages combined. Using the IVW method for analysis, results showed a significant Mendelian causal relationship (p < 0.05) between 36 types of food and BMD. Specifically, low-fat hard cheese and other milk-based puddings had a negative impact on BMD in the 45–60 years group and across all ages. Meanwhile, sprouts positively affected BMD in the over 60 years group and across all ages. The remaining 33 foods had significant impacts only within specific age groups: In the 0–15 years group, increased intake of crumble, doughnuts, breaded fish, avocado, and poultry was associated with higher BMD, beneficial for bone growth; green tea slightly decreased BMD, while flavoured milk had a negative correlation with BMD. For the 15–30 years group, soya dessert, tofu, sweet potato, milk-based pudding, dark chocolate, other sweets, rose wine, and pizza were positively correlated with BMD; milk, sponge pudding, white fish, and vegetarian alternatives negatively correlated with BMD. In the 30–45 years group, pineapple, omelette, and green beans positively impacted BMD; sweets and white rice negatively affected BMD. For the 45–60 years group, besides low-fat hard cheese and other milk-based puddings, cake, tinned tomato, and squash consumption reduced BMD; moderate intake of soft cheese and cappuccino helped maintain BMD. In the over 60 years group, sprouts intake promoted BMD, whereas mozzarella consumption negatively correlated with BMD. Across all ages, sprouts and sausages positively influenced BMD; standard tea slightly promoted BMD; carrots, low-fat hard cheese, and other milk-based puddings were negatively associated with BMD.

We adopted the food classification method from the literature22, categorizing foods into three major categories (plant-based foods, animal-based foods, and mixed foods), and further subdividing them into seven subcategories, including carbohydrates, fruits and vegetables, soy products, beverages, meat and eggs, dairy products, and desserts (Table 3). Browsing the IVW (Osteo) webpage in the ONMR database, we analyzed and generated MR results forest plots for six different age groups (Fig. 3) to investigate the trends of how dietary intake affected BMD across the lifespan. The results showed that during adolescence (0–15 years) and early adulthood (15–30 years), dietary patterns were predominantly associated with increasing BMD, with approximately 70% and 67% of foods exhibiting positive effects on BMD, respectively. As age progressed, this proportion gradually decreased: during middle age (30–45 years), about 60% of foods still contributed to maintaining BMD. However, in late middle age (45–60 years) and old age (over 60 years), this proportion significantly dropped to 29% and 50%, respectively, indicating a shift from promotion to inhibition in the impact of diet on BMD. Notably, the intake of carbohydrate-related nutrients (such as rice, sugars, and puddings) might become an important risk factor for BMD decrease after middle age.

Fig. 3: The effect of food intake on bone mineral density (BMD) in different age groups.
figure 3

The figure details the impact of food intake on BMD across different age groups (0–15 years, 15–30 years, 30–45 years, 45–60 years, over 60 years, and all ages combined). Red lines in the figure indicate an association between food intake and decreased BMD, while blue lines denote an association between food intake and increased BMD.

Full size image
Table 3 Statistical descriptions of the intake food and categories
Full size table

Search, visualization, and download functions of the ONMR database

The ONMR database provided convenient search, plotting, and download functions, fulfilling users’ needs from data acquisition to visualization. Upon accessing the homepage of the database, users could use the quick search function to locate the required information (as shown in Fig. 4A). The specific search process includes: entering the target food name (clicking the search button without entering any text will allow you to enter the filtering mode), clicking the “Search Nutri” button, and then selecting the appropriate record on the redirected page. The search results were presented in two structured tables. The first table displayed the MR analysis results between food intake and all orthopedic-related terms, while the second table provided detailed MR results for each single SNP.

Fig. 4: A concise user guide to the ONMR database.
figure 4

A Access and retrieve data from the database; B Visualize the results of MR calculations; C Download raw (or analyzed) data.

Full size image

Based on the search results, users could utilize the database’s built-in plotting functions for visual analysis (as shown in Fig. 4B). The plotting methods were divided into two types: the first involved clicking the “Visualization Link” button after completing the search, which generated a one-to-one MR forest plot for single food intake versus a single disease from the first table. The second method involved using the “Browse” function in the navigation bar to select either the IVW (Osteo) or IVW (Nutri) method, which allowed users to create many-to-one or one-to-many MR forest plots. These plots respectively displayed analysis results related to diseases or food intake.

The database also supported two modes of data download functionality, allowing users to save and analyze the data (as shown in Fig. 4C). The first mode was the search (plot) download mode, where users could click the “download csv” button on the search results page to download the current table information, or on the plotting page, click “download png” to download image files and “download json” to download the raw data used for plotting. The second mode was accessible through the “download” page in the navigation bar, which provided downloads of all raw GWAS annotation information, including 210 food intake entries, 503 bone-related terms, and 6,346 significant SNP entries.

Discussion

The dynamic interplay between dietary components and skeletal health spans critical biological processes from osteogenesis to age-related bone loss. In this study, we present ONMR, the largest platform of Mendelian randomization analysis, which integrating 210 nutrient items with 503 skeletal phenotypes to provides unprecedented insights into the causal relationships between food chemistry and orthopedic pathophysiology through over 100,000 MR associations.

The ONMR database not only facilitates the exploration the multifaceted impacts of dietary intake on orthopedic disorders, but also supports diverse research in the field of food science, such as integrating dietary structure, food culture, and food consumption habits to deeply analyze the mechanisms underlying orthopedic diseases. Taking food culture (e.g., regional differences) as an example, according to records in the ONMR database, the intake of shrimp/crab, shellfish, and oily fish might help increase bone mineral density. These foods fell under the broader category of seafood products, which were widely consumed in coastal regions23. A multi-center study examining the dietary patterns of coastal residents demonstrated that the intake of oily fish had reduced the incidence of osteoporosis and osteopenia in certain populations24. As one of the key quantitative indicators of osteoporosis23,25, changes in BMD suggested that specific dietary habits might have protective effects on bone health and potentially slow age-related bone loss. For instance, oily fish were rich in ω-3 fatty acids and vitamin D, two nutrients proven to enhance BMD by modulating inflammatory responses, promoting calcium absorption, and improving bone metabolism26. The impact of diet on BMD varies depending on food types and age groups. In the “Age-dependent relationship between diet and bone mineral density: from promotion to suppression” section, the analysis based on bone density, intake of rice and carbohydrates after the age of 30 would lead to a decline in bone density, likely due to age-related changes in human metabolic levels. This suggested the importance of adjusting dietary structure according to age, such as reducing carbohydrate intake. Although the dietary patterns of coastal regions provided potential nutritional intervention strategies for preventing osteoporosis, the exact mechanisms by which these foods positively affected bone density require further research and validation, particularly regarding their applicability across different populations and their long-term effects.

During the exploration of the ONMR database, we also identified several interesting and noteworthy associations. For example, there appeared to be a potential link between bread consumption and a reduced risk of upper arm and periarthritis, including decreased probabilities of wrist and hand vascular injuries, shoulder region lesions, calcific tendinitis of the shoulder, and other shoulder-related disorders. To validated this interesting phenomenon, we used the keywords “bread” and “shoulder” searched in the PubMed. The results indicated that the simple everyday act of slicing bread is often used as part of shoulder rehabilitation exercises27. This suggested that the repetitive motion of slicing bread, associated with moderate bread consumption, could play a positive role in the prevention of shoulder-related diseases.

Previous study reported that coffee consumption may be associated with an increased risk of osteoarthritis20,28,29. In our study, among the 17 osteoarthritis-related GWAS datasets included in our database, 14 showed a positive association between coffee intake and osteoarthritis risk, with effect sizes (β) ranging from 0.0118 to 0.7737. Based on this observation, we conducted a systematic annotation across these 17 datasets and identified five food items — Cereal, Cheese, Coffee, Dried fruit, and Tea — that demonstrated potential causal associations with osteoarthritis in at least 10 out of the 17 datasets (Supplementary Table 2). Among these, Cereal and Cheese were predominantly linked to protective effects against osteoarthritis, with β values ranging from −0.6872 to −0.0207 for Cereal and from −0.5623 to −0.0140 for Cheese. In contrast, both Coffee and Tea consistently showed risk-enhancing effects, with β values for Tea ranging from 0.0147 to 0.4288. Notably, Dried fruit exhibited heterogeneous effects: it was associated with an increased risk of hip osteoarthritis (dataset ID: ieu-a-1169), yet showed protective effects in the other 10 datasets (β: −0.8640 to −0.0173). These findings highlight the importance of integrating multiple datasets to obtain a more comprehensive and robust understanding of the potential causal relationships between dietary intake and osteoarthritis risk.

Food intake was closely related to human health, and bone research spanned the entire life course, encompassing multiple important areas such as bone development, bone health, bone aging, and bone degeneration. In exploring the impact of food intake on orthopedic health, MR analysis served as a powerful tool to uncover potential causal relationships between different factors. Large-scale MR studies focusing on the relationship between food intake and orthopedic health remained relatively scarce. We believed that the ONMR database would support researchers in quickly reviewing the effects of food intake on orthopedic disorders and also enabled users to conduct more detailed investigations into specific foods or bone-related topics based on their individual research interests.

Despite its comprehensive scope, the ONMR has several limitations. First of all, the GWAS data used in ONMR were derived from European-ancestry cohorts such as UK Biobank, FinnGen and EBI. While these datasets offer large sample sizes for data mining, the generalizability of these datasets was limited and had biases that prevented them from being universally studied for other populations. In the future, more data sources can be included, and the studies may be conducted on different continents to enrich the applicability of the results. Also, the dietary exposures were defined using broad intake categories without granularity in preparation methods, frequency or synergistic nutrient interactions. This simplification might be addressed in the future with more precise digital measurements. Additionally, the bias caused by the weak instruments requires further consideration. Even though most of the instrumental variables (IVs) were robust (with F > 10), weak instruments for certain nutrients may affect the stability of the results. In the future, with the accumulation of more valuable data, study design will refine dietary phenotyping and integrate pleiotropy-robust MR methods.

In our two-sample MR design, exposure data (dietary intake) were exclusively sourced from the UKB, while outcome data (orthopedic disorders) including 41.36% (208/503) UKB derived datasets and 21.27% (107/503) among which derived from MRC-IEU datasets. The theoretical sample overlap rate was estimated as 2.98% based on the median sample sizes (exposure median of 64,949 and outcome median of 463,010). To quantify the potential bias, we conducted simulations across 483 exposure-outcome pairs under three levels of overlap: 0.03, 0.085 (the mean of the two extremes), and 0.14. Results indicated a systematic negative bias (mean: −3.5%) in causal estimates (β), implying that true effect sizes might be underestimated in our analyses. This aligns with prior reports that participant overlap can inflate instrument strength and attenuate effect estimates30. While this bias is modest overall, future iterations of ONMR will integrate overlap-robust MR methods and diverse population cohorts to mitigate such limitations.

In summary, we present the ONMR database, a comprehensive resource designed to elucidate causal relationships between dietary intake and orthopedic diseases through large-scale Mendelian randomization (MR) analysis. By systematically integrating GWAS data of 210 nutrient items and 503 skeletal phenotypes, ONMR services the purpose of link the relations between dietary effects on skeletal diseases including but not limited to protective, detrimental, age-dependent relationships. This platform enables dynamic querying, visualization, and personalized analysis, bridging nutritional science and orthopedics to advance precision medicine. Furthermore, ONMR provides valuable insights into the critical role of diet in skeletal health and establishes a robust framework for interdisciplinary research, supporting the development of preventive and personalized nutritional interventions. While this resource represents a significant step forward, we note that the current scope of orthopedic terms requires further expansion. Future work will incorporate additional disease classifications through deeper integration with clinical orthopedic practice, followed by continuous database refinement and updates.

Methods

Data source

The raw GWAS data primarily originated from UK Biobank (https://www.ukbiobank.ac.uk), FinnGen (https://www.finngen.fi/en), EBI (https://www.ebi.ac.uk/gwas/), and several other publications/databases. During the retrieval of genome-wide association study (GWAS) data related to dietary intake, we conducted an initial search using the keyword “intake,” yielding a total of 610 records. We then removed duplicates based on GWAS-ID. Subsequently, we excluded data containing word such as “Frequency,” “Average,” “Versus,” “Added,” “Diagnoses,” and “Symptoms.” Additionally, we excluded one term related to chemical molecule intake, namely “Bap intake.” After the aforementioned filtering steps, this study obtained 536 valid GWAS records originating from three distinct datasets, covering 210 types of food. Among these, the UK Biobank datasets, as the largest analysis cohort, fully encompassed intake records for all 210 foods. To ensure consistency and reliability of the data in subsequent nutritional analyses, this study ultimately selected the GWAS records corresponding to these 210 foods from the UK Biobank datasets, which were used as the exposure variables for MR analysis in our database. Simultaneously, this study used the following terms as search keywords: Osteo, Bone, Neck, Shoulder, Elbow, Wrist, Knee, Ankle, Joint, Tendon, Lumbar, Vertebra; Arthritis, Bursitis, Spondylitis, Synovitis, Tendinitis; Chondromalacia, Fracture, Hyperostosis, Sciatica, Spondylosis. The records were also deduplicated based on GWAS-ID, resulting in 503 GWAS records related to bone terms, which were used as the outcome variables for MR analysis.

Mendelian randomization analysis

This study employed the TwoSampleMR package (version 0.6.6)31 in R for MR analysis. During the calculation of exposure factors (dietary intake), a p1 value of 5e-6 was set to screen SNPs associated with the exposure, and the clustering option was enabled to remove SNPs in high linkage disequilibrium (R-squared=0.001 and kb=10000). In the analysis of outcome factors (osteo-related terms), we extracted SNPs consistent with the exposure factors from GWAS data, ensuring high correlation between SNPs and their proxy SNPs by setting proxies=TRUE and rsq=0.8. Alleles were aligned using align_alleles=1, and palindromic SNPs were handled using the palindromes=1 strategy. A minor allele frequency threshold (maf_threshold=0.3) was applied to ensure the quality and applicability of the selected SNPs.

Website architecture

Our database architecture is based on the Linux operating system (CentOS 7.9.2009) and uses the Apache HTTP server (version 2.4.6) to ensure system stability and efficiency. Front-end development employs HTML, CSS, and JavaScript to provide users with an intuitive and highly interactive interface experience. The back-end is built using the Python language (version 3.12.8) combined with the Flask framework (version 3.0.3), seamlessly integrating with a MySQL database through the Flask-SQLAlchemy extension (version 3.1.1) to support complex data operation requirements. Additionally, SQLAlchemy (version 2.0.37) is adopted as the ORM tool to enhance code maintainability and flexibility of the data access layer. During development, SQLite (version 3.45.3) is used to simplify the local testing process. Furthermore, our technology stack includes the Jinja2 template engine (version 3.1.4) for dynamically generating HTML pages.

Interactive visualization and security

For database visualization, a variety of front-end technologies are employed to enhance user experience. For instance, DataTables.js (version 1.10.15) and D3.js (version 6.7.0) are utilized to achieve dynamic presentation and interactive functionalities of the data, ensuring that users can efficiently and intuitively explore complex datasets. To support these functionalities, the database integrates modules such as jQuery (version 1.12.4), the jQuery DataTables plugin (version 1.10.15), and the DataTables Buttons extension (version 2.3.6), providing users with a rich operational interface and export options. Furthermore, to ensure security during data transmission, access to the database is encrypted using the HTTPS protocol.