SALIVA+: a FAIR database providing data on diet-related contaminants in saliva

The SALIVA+ FAIR database provides curated quantitative data on salivary concentrations of diet-related contaminants, supporting the development of saliva as a non-invasive matrix for human biomonitoring. The database includes 11 groups of contaminants, namely, nitrite and nitrate, heavy metals, bisphenols, polycyclic aromatic hydrocarbons (PAHs), biogenic amines, pesticides, advanced glycation end products (AGEs), perchlorate, microplastics, parabens, and phthalates, extracted from peer-reviewed studies focused on saliva biomonitoring. The database was developed through a combined approach of advanced text mining and manual validation. An initial search was performed using the term “saliva” in the PubMed Entrez API and the Scopus database, retrieving approximately 58,630 and 61,330 articles, respectively. Programmatic filters using a dictionary of over 500 target diet-related contaminants were applied, retaining 6,380 records (search period between 01/01/2000 and 18/01/2025). After removing duplicates and applying exclusion criteria, 1,687 articles were manually screened. Only 104 studies were identified with extractable quantitative data relevant to saliva-based biomonitoring. The database includes harmonized metadata on salivary concentration ranges, analytical methods, population characteristics, exposure pathways, and potential correlations with systemic biomarkers. Many contaminant classes, such as mycotoxins, HAAs, PFAS, dioxins, furans, brominated flame retardants, and UV filters, were not quantitatively assessed in saliva. Nevertheless, the included compounds provide a critical foundation for evaluating the suitability of saliva as a non-invasive matrix for dietary exposome research. By consolidating heterogeneous findings into a standardized FAIR database, SALIVA+ supports exposure modelling, method development, and non-invasive biomonitoring strategies, particularly valuable in studies involving children or vulnerable populations.SALIVA+ FAIR数据库收录经整理审核的膳食相关污染物唾液浓度定量数据，为确立唾液作为人体生物监测非侵入性基质的应用路径提供支撑。该数据库涵盖11大类污染物，分别为亚硝酸盐与硝酸盐、重金属、双酚类化合物、多环芳烃（Polycyclic Aromatic Hydrocarbons, PAHs）、生物胺、农药、晚期糖基化终末产物（Advanced Glycation End Products, AGEs）、高氯酸盐、微塑料、对羟基苯甲酸酯类以及邻苯二甲酸酯类，所有数据均提取自聚焦唾液生物监测的同行评议研究。 本数据库通过先进文本挖掘与人工核验相结合的方式构建。首先于PubMed Entrez API及Scopus数据库中以关键词"saliva"进行检索，分别获取约58630篇与61330篇文献。随后通过内置包含500余种目标膳食相关污染物的词典进行程序化筛选，保留了6380条记录（检索时间范围为2000年1月1日至2025年1月18日）。在去除重复文献并应用排除标准后，对1687篇文献开展人工筛查，最终仅筛选出104项可提取唾液生物监测相关定量数据的研究。 该数据库包含标准化元数据，涵盖唾液浓度范围、分析方法、人群特征、暴露途径以及与系统性生物标志物的潜在关联。已有研究未对真菌毒素、杂环胺（HAAs）、全氟和多氟烷基物质（PFAS）、二噁英、呋喃类、溴化阻燃剂以及紫外线过滤剂等诸多污染物类别开展唾液定量评估。但纳入本数据库的化合物仍为评估唾液作为膳食暴露组研究中非侵入性基质的适用性提供了关键基础。 本数据库将异质性研究结果整合为符合FAIR（可发现、可访问、可互操作、可复用）标准的标准化数据库，可为暴露建模、分析方法开发以及非侵入性生物监测策略提供支持，在涉及儿童或脆弱人群的研究中尤为具有应用价值。