112

6,850 千元

2023/04/17

2023/12/31

曾宇鳳

國環院氣候變遷研究中心

林志鴻

國立臺灣大學

環境監測及檢測

綠色化學、危害資料庫、危害評估預測

Green chemistry, Database for hazardous chemical, Hazard assessment prediction

郭天爵

02-33664888#403

cot@cmdm.csie.ntu.edu.tw

國家環境研究院「綠色消費化學品毒性預測技術建立」計畫著重於環境檢體分析，從儀器所偵測得之新興綠色消費化學物進行化學物毒性預測；化學物質管理署「化學物質綠色替代診斷模組建置及決策支援計畫(1/4)」計畫著重於建立高風險化學物質之安全替代化學物質搜尋、評估及篩選之系統，以高風險化學品監管評估層面建立一套系統可供化學署與產學界快速搜尋取得其危害資訊與風險等級，並提供可參考之安全替代物。本計畫所建置之系統可提供化學署搜尋化學物質的危害評估，作為參考。 　　首先，計畫蒐集大量關於化學品毒性的相關資料，包括高關注或各國監管數據、研究報告和相關文獻。這些資料用於建立毒性預測模型，以識別和分析化學品的關鍵特徵和相關性。這些特徵成為預測模型的基礎，並用於預測新興化學品的毒性。為了驗證團隊之預測模型的準確性和可靠性，團隊使用已有的測試資料進行驗證。這些資料將與我們的預測結果進行比較，以評估模型的準確度。團隊也與相關機構和專家進行會議，諮詢專家建議，進一步優化系統之預測技術。 　　本計畫透過國內外資訊庫蒐集整合，結合建立定量結構活性關係(Quantitative Structure-Activity Relationship, QSAR)模型，串接分析儀器，以建立綠色消費化學品毒性預測作業流程，申請臺灣與美國暫時性專利，並辦理資料搜尋技術之教育訓練，落實綠色化學的推動。

綠色化學、危害資料庫、危害評估預測

112BB003_for_public.pdf

19MB

2024/2/16 上午 12:00:00

National Environmental Research Academy’s “Establishment of Green Consumer Chemical Toxicity Prediction Technology” project focuses on environmental specimen analysis. It involves predicting the toxicity of emerging green consumer chemicals detected by instruments. On the other hand, Chemicals Administration’s “Development of a Green Chemical Substitution Diagnostic Module and Decision Support Plan(1/4)” project emphasizes establishing a system for searching, assessing, and screening safe alternative chemicals for high-risk chemicals. This system is designed to provide a rapid means for the Chemicals Administration and the industry to search for and obtain hazard information and risk levels for high-risk chemicals on the regulatory assessment front. Additionally, it offers alternative safe options. The system developed by this project can also assist the Chemicals Administration in searching for hazard assessments of chemical substances. 　　First, we will collect a large amount of relevant data on chemical toxicity, including existing experimental data, research reports, and literature. This data will be used to establish a comprehensive and reliable toxicity prediction model. We will employ machine learning and statistical analysis techniques to process and mine the data, identifying key features and correlations of chemicals. These features will form the basis of the prediction model and will be used to predict the toxicity potential of new chemicals. 　　To validate the accuracy and reliability of our prediction model, we will conduct validation using our collected data. The laboratory data will be compared with our prediction results to assess the model’s accuracy. We will also collaborate with relevant organizations and experts to gather their opinions and suggestions, further improving and optimizing our prediction technology. Lastly, we will establish an open and sustainable platform to provide chemical toxicity prediction services. Through this platform, consumers can access toxicity information on different chemicals, based on their individual needs and health considerations. By integrating data from domestic and international databases, implementing quantitative structure-activity relationship (QSAR) models, and connecting with analytical instruments, we will establish a workflow for green consumer chemical toxicity prediction. Temporary patents will be sought in Taiwan and the United States, and educational training on data searching techniques will be conducted to promote the adoption of green chemistry. Overall, this project aims to bridge the gap between green consumerism and chemical toxicity prediction, ensuring a sustainable approach to chemical product selection and usage.

Green chemistry, Database for hazardous chemical, Hazard assessment prediction