http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
성낙도 한국응용생명화학회 2003 Applied Biological Chemistry (Appl Biol Chem) Vol.46 No.3
Biological Hammett Equation에 기초하여 Hansch-Fujita식으로 제안된 정량적인 구조 활성상관(QSAR) 기법(Sung, Nack-Do (2002) Development of new agrochemicals by quantitative structure-activity relationship(QSAR) methodology. Kor. J. Pestic. Sci. 6: 166-174, 231-243 및 7: 1-11)에 따른 새로운 농약의 탐색과 개발에 관하여 1990년도를 전후한 국내에서 이루어진 QSAR 기법 중 주로 2D QSAR 기법의 활용연구 현황에 대하여 조명하였다. It was reviewed for the status of domestic research before and after 1990's for search of a new pesticides using 2D QSAR of quantitative structure-activity relationship (QSAR) methodologies (Sung, Nack-Do (2002) Development of new agrochemicals by quantitative structure-activity relationship (QSAR) methodology. Kor. J. Pestic. Sci. 6, 166-174, 231-243 & 7, 1-11) which was proposed according to Hansch-Fujita equation based on the concept of biological Hammett equation.
Bedadurge, Ajay B.,Shaikh, Anwar R. Korean Chemical Society 2013 대한화학회지 Vol.57 No.6
Quantitative structure-activity relationship (QSAR) analysis for recently synthesized imidazole-(benz)azole and imidazole - piperazine derivatives was studied for their anticancer activities against breast (MCF-7) cell lines. The statistically significant 2D-QSAR models ($r^2=0.8901$; $q^2=0.8130$; F test = 36.4635; $r^2$ se = 0.1696; $q^2$ se = 0.12212; pred_$r^2=0.4229$; pred_$r^2$ se = 0.4606 and $r^2=0.8763$; $q^2=0.7617$; F test = 31.8737; $r^2$ se = 0.1951; $q^2$ se = 0.2708; pred_$r^2=0.4386$; pred_$r^2$ se = 0.3950) were developed using molecular design suite (VLifeMDS 4.2). The study was performed with 18 compounds (data set) using random selection and manual selection methods used for the division of the data set into training and test set. Multiple linear regression (MLR) methodology with stepwise (SW) forward-backward variable selection method was used for building the QSAR models. The results of the 2D-QSAR models were further compared with 3D-QSAR models generated by kNN-MFA, (k-Nearest Neighbor Molecular Field Analysis) investigating the substitutional requirements for the favorable anticancer activity. The results derived may be useful in further designing novel imidazole-(benz)azole and imidazole-piperazine derivatives against breast (MCF-7) cell lines prior to synthesis.
최현철,하시영,임우석,양재경 경상국립대학교 농업생명과학연구원 2024 농업생명과학연구 Vol.58 No.1
Plants synthesize antioxidant compounds as a defense mechanism against reactive oxygen species. Recently, plant-derived antioxidantcompounds have attracted attention due to the increasing consumer awareness in the heath industry. However, traditional methods formeasuring the antioxidant activity of these compounds are time-consuming and costly. Therefore, our study constructed a quantitativestructure-activity relationship (QSAR) model that can predict antioxidant activity using graph convolutional networks (GCN) from plantstructural data. The accuracy (Acc) of the model reached 0.6 and the loss reached 0.03. Although with lower accuracy than previouslyreported QSAR models, our model showed the possibility of predicting DPPH antioxidant activity in a wide range of plant compounds(phenolics, polyphenols, vitamins, etc.) based on their graph structure.
Compounds Toxicity Prediction Using Convolution Neural Network Model
Waqar Ahmad,Kil To Chong 제어로봇시스템학회 2021 제어로봇시스템학회 각 지부별 자료집 Vol.2021 No.12
The compound toxicity is its ability to cause damaging effects on single cell or cells group or organ of the body. Due to bioassay advancements in recent years, new drugs emerged day by day and hence chemical toxicity data also increased. Moreover, traditional toxicity analysis methods failed to process large amount of toxicity data. Using these large amount of toxicity data, deep learning methods are useful for building Quantitative structure-activity relationship (QSAR) models for toxicity prediction. We used the convolution neural network model for toxicity prediction using SMILES images. This method achieved the 79% F1-score for toxic and non-toxic predictions. We hope that out model will contribute towards toxicity prediction in de novo drug discovery.
Hyun-Joo Chang,Eun Hye Choi,Hyang Sook Chun 한국식품과학회 2008 Food Science and Biotechnology Vol.17 No.3
The quantitative structure-activity relationships (QSAR) study of antioxidative anthocyanidins and their glycosides were evaluated using 4 different assays of Trolox equivalent antioxidant capacity (TEAC), superoxide radical (O<sub>2</sub> <sup>?</sup>), hydrogen peroxide (H<sub>2</sub>O<sub>2</sub>), and peroxynitrite radical (ONOO<sup>?</sub>) scavenging with TSAR software. Four models were developed with significant predictive values (r<sup>2</sup> and p value), which indicated that the antioxidant activities were mainly governed by the 3-dimensional structural energy (torsional energy), constitutional properties (the number of hydroxyl and methyl groups), and electrostatic properties (heat of formation, and dipole, quadrupole, and octupole components). This QSAR approach could contribute to a better understanding of structural properties of anthocyanidins and their glycosides that are responsible for their antioxidant activities. It might also be useful in predicting the antioxidant activities of other anthocyanins.
Mehdi Khoshneviszadeh,Amirhossein Sakhteman 한국응용생명화학회 2016 Applied Biological Chemistry (Appl Biol Chem) Vol.59 No.3
Exploring predictive QSAR models for dopamine catechol structures could be used in designing more potent ligands. In this study, efforts were taken to find out the most important molecular features responsible for the biological activity of catechol structures. All 2D descriptors of Dragon including constitutional, topological, molecular walk counts, BCUT descriptors, Galvez topological, 2D autocorrelations, functional groups, atom-centred fragments, empirical descriptors and properties were calculated for the structures. Two non-linear modelling methods (PC-LS-SVM and PC-ANFIS) were used and compared in this QSAR study. The results revealed the more predictive ability of PC-LS-SVM in the QSAR analysis of the compounds with catechol substructure. The roles of topological properties and number of hydrogen bond donors group as molecular features responsible for the activity of the compounds were discussed. The obtained QSAR models can be used in future studies of drug development for human dopamine D2 receptor.
Isoflavones as modulators of adenosine monophosphate-activated protein kinase
정혜령,안승현,김범수,신순영,이영한,임융호 한국응용생명화학회 2016 Applied Biological Chemistry (Appl Biol Chem) Vol.59 No.2
Adenosine monophosphate-activated protein kinase (AMPK) is expressed in all eukaryotic cells and can therefore be found in vertebrates, invertebrates, and plants. Since AMPK participates in the regulation of homeostasis on various levels, small compounds that can modulate AMPK activity could be valuable research tools. Several flavonoids can modulate AMPK. Here we investigated the modulatory effect of 37 isoflavones on AMPK activity using an in vitro kinase assay. Because the relationship between the structural properties of flavonoids and their modulatory activities has not been elucidated yet, we used comparative molecular field analysis to derive the structural conditions for modulation of AMPK activity. The molecular binding mode of isoflavones to AMPK was elucidated using in silico docking studies. The findings presented here can aid in the design of new modulators with better specificity for AMPK.
Seketoulie Keretsu,Pavithra Kuruchi Balasubramanian,Swapnil Pandurang Bhujbal,조승주 대한화학회 2017 Bulletin of the Korean Chemical Society Vol.38 No.11
Cyclin-dependent kinase 2 (CDK2) plays important roles in cell cycle regulation. Owing to its multiple roles in many cancer types, it is considered as a significant target for cancer drug design. In this study, we used docking techniques and 3D-quantitative structure–activity relationship (3D-QSAR) studies on a series of 6-substituted 2-arylaminopurine derivatives as CDK2 kinase inhibitors. Receptor-guided comparative molecular field analysis (CoMFA) (q2 = 0.653, optimal number of component [ONC] = 6, r2 = 0.965) and comparative molecular similarity indices analysis (CoMSIA) (q2 = 0.718, ONC = 6, r2 = 0.872) models were developed. Validation by progressive scrambling, bootstrapping (BS), and leave-five-out method suggests that the developed 3D-QSAR models (CoMFA and CoMSIA) have reasonable predictive ability and reliability. Docking analysis revealed five hydrogen bond interactions between the compound 13 (most active compound) and CDK2 active site residues namely Glu81, Leu83, and Asp86. Contour map analysis gave comprehensive information regarding favorable and unfavorable substitution groups. Bulky, electropositive, and hydrophobic substitution groups at the R2 region can enhance the inhibitory activity of compounds whereas electronegative, hydrogen bond donor, and nonhydrophobic substitution groups at the position R1 will increase the inhibitory activity. Our contour map results can be used as a guideline in designing new potent compounds for CDK2 inhibition.
Ying Cui,Qinggang Chen,Yaxiao Li,Ling Tang 대한약학회 2017 Archives of Pharmacal Research Vol.40 No.2
Flavonoids exhibit a high affinity for the purifiedcytosolic NBD (C-terminal nucleotide-binding domain) ofP-glycoprotein (P-gp). To explore the affinity of flavonoidsfor P-gp, quantitative structure–activity relationship(QSAR) models were developed using support vectormachines (SVMs). A novel method coupling a modifiedparticle swarm optimization algorithm with randommutation strategy and a genetic algorithm coupled withSVM was proposed to simultaneously optimize the kernelparameters of SVM and determine the subset of optimizedfeatures for the first time. Using DRAGON descriptors torepresent compounds for QSAR, three subsets (training,prediction and external validation set) derived from thedataset were employed to investigate QSAR. Withexcluding of the outlier, the correlation coefficient (R2) ofthe whole training set (training and prediction) was 0.924,and the R2 of the external validation set was 0.941. Theroot-mean-square error (RMSE) of the whole training setwas 0.0588; the RMSE of the cross-validation of theexternal validation set was 0.0443. The mean Q2 value ofleave-many-out cross-validation was 0.824. With moreinformations from results of randomization analysis andapplicability domain, the proposed model is of good predictiveability, stability.