http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Review of toxicity prediction studies using toxicity database
Jaeseong Jeong(정재성),Jinhee Choi(최진희) 환경독성보건학회 2021 한국독성학회 심포지움 및 학술발표회 Vol.2021 No.5
Recently, computational toxicology has emerged that predicts toxicity without conducting toxicity tests at all. This has become possible due to the rapid development of computer technology, and various computational toxicology techniques such as quantitative structure-activity relationship (QSAR) that predict toxicity based on the structure of chemical substances are attracting attention. Currently, research are underway to apply artificial intelligence techniques used to process big data in various fields to toxicology, mainly in scientifically advanced countries. The competition for the development of toxicity prediction models using artificial intelligence is accelerating, and techniques are becoming increasingly complex. To develop a toxicity prediction model using artificial intelligence and use it for regulation, it is necessary to understand the recent development. In this study, we analyze toxicity prediction studies using artificial intelligence techniques, and summarize artificial intelligence algorithms and prediction performance used in recent papers. We have analyzed over 70 papers published since 2014. Models have been developed to predict about 30 different toxicity endpoints using more than 20 toxicity databases. In the development of the model, MACCS fingerprint and random forest algorithms were used the most. The use of artificial intelligence techniques in the development of toxicity prediction models is a fairly new challenge, requiring active and diverse efforts toward a scientific accord and regulatory application. The comprehensive overview provided in this study could be used as a useful guide to further development and application of toxicity prediction models.
화학물질 독성고속대량스크리닝 프로그램 ToxCast<sup>TM</sup> 분석
정재성 ( Jaeseong Jeong ),임창원 ( Changwon Lim ),정다운 ( Da Woon Jung ),최진희 ( Jinhee Choi ) 한국환경분석학회 2019 환경분석과 독성보건 Vol.22 No.2
Asthenumberofnewchemicalsincreases,traditionalanimaltestinghaslimitationsinevaluatingthetoxicityofchemicals.InEuropeandtheUnitedStates,theuseofnon-animalalternativetoxicitytestmethodsforchemicalregulationisencouragedandrelatedresearchisactivelycarriedout.IntheUnitedStates,theToxCastprogramisinprogressusinghigh-throughputscreening(HTS)techniques.TheprogramproducedHTSinvitrotestresultsfor1200assaysand9000chemicalstodevelopchemicaltoxicitypredictionmodelsandsettheprioritiesfortoxicityassessment.Inthisreview,weexploredtheoutlineofToxCast,thetypesofassaysused,andtheprogressofeachphase.WealsoexploredhowtoanalyzethenumeroustoxicitydatageneratedthroughtheToxCastprogram,includingpre-processing,dose-responsemodelfit,conclusionandcategorization.Finally,implicationsfordomesticutilizationoftheToxCastdatabasearederived.