Construction Of A Database Of Per-and Polyfluoroalkyl Substances Loads In Organisms And Data Mining

Per-and polyfluoroalkyl substances(PFASs)are a group of synthetic industrial chemicals whose molecules consist of methyl or methylene groups fully substituted with fluorine molecules.PFASs are widely present in the environment and organisms,and its ecological risks and health effects have received widespread attention.Current PFASs loads data in organisms are fragmented and disorganized,failing to provide systematic knowledge.Therefore,this paper leverages database technology to integrate PFASs loads data in organisms,constructing a database for PFASs load in organisms and aiming to analyze and mine the database.The main research contents and results are as follows:A database PFASs loads in organisms was constructed and organized to accommodate multidimensional biomonitoring data including time series,spatial coordinates,and species information.The database architecture comprises four datasets: biological pollution dataset,data source dataset,species information dataset,and pollutant information datasets.We developed a method of rapid database generation based on modular collection of biological monitoring data and completed the preliminary establishment of the database.As of April 2023,the database has collected 27284 data records between 1966 and 2019,comprising 563 sampling points across all global regions,including ocean,lake,river,and land habitats.The recorded data includes 79 PFASs,such as PFCAs,PFSAs,and 430 species,such as Actinopterygii,Aves,and Mammalia.This paper narrows the focus and examines bioload data from nine long-chain PFASs: PFHx S,PFOS,PFOA,PFNA,PFDA,PFUn DA,PFDo DA,PFTr DA,and PFTe DA.Machine learning is employed to interpolate and fill missing data,improve tissue conversion coefficient calculations to resolve variability in biological data,and assess the health risks of PFASs resulting from aquatic product consumption using the target hazard quotient.The paper identified that:(1)The median concentration of PFASs in organisms globally is 23.29 ng/g(95% CI=0.59-1063.58 ng/g),and the large confidence interval reflects biomonitoring data variability.(2)The Western Hemisphere has significantly higher PFASs load levels in organisms than the Eastern Hemisphere,with median concentrations of55.85 ng/g and 14.41 ng/g,respectively,indicating that PFASs emissions may primarily occur in the Western Hemisphere.(3)Median concentrations of PFASs in terrestrial organisms are 88.33 ng/g,while those in aquatic organisms,including marine,lake,and river habitats,are11.71 ng/g,14.89 ng/g,and 11.66 ng/g,respectively.This difference may be due to differences in respiratory pathways.(4)PFOS contribution accounts for the majority of PFASs load in Actinopterygii,Aves,and Mammalia with contribution rates of 55.16%,59.48%,and 51.99%,respectively.(5)PFASs load levels in organisms are generally decreasing over time.In 2019,the total median concentration of the nine-long chain PFASs in organisms was 99.77% lower than the amount in 2002.(6)The regions with health risks globally mainly include East Asia,North America,and Western Europe.Freshwater fish is the water product with the highest probability of causing health risks.The main health risks come from PFOA and PFOS.In summary,this paper developed a database of PFASs loads data in organisms and introduces new methods and tools to facilitate the database development.Moreover,the paper analyzes and investigates the data of PFASs loads in organisms,Spatiotemporal distribution,and health risks associated with PFASs in organisms.