Many bioactive peptides have been reported from various food proteins through the conventional activity-guided-purification approach; however, the rationale behind the selection of conditions for the production of the bioactive peptides has not been extensively explored. The purposes of the study were to provide the rationale behind the selection of conditions, and to develop an innovative strategy to explore the most potent peptides within egg proteins through an integrated QSAR and bioinformatics approach. Thermolysin-pepsin hydrolysate of ovotransferrin was predicted as the best condition for production of ACE-inhibitory peptides. Three predicted peptides, IRW, LKP and IQW, were successfully released from ovotransferrin. Simulated gastrointestinal incubation showed IQW was stable while IRW and LKP can be degraded into dipeptides (IR and KP respectively). Peptides produced from the study will have the potential to be developed as functional foods and nutraceuticals for the prevention of hypertension, a disease affecting ~ 31% of the adult population. / Food Science and Technology

Virtual screening is a fast, low cost method to identify potential small molecule therapeutics from large chemical databases for the vast amount of target proteins emerging from the life sciences and bioinformatics. In this work, we applied several conventional and newly developed virtual screening approaches to identify novel non-steroidal ligands for the human and zebrafish sex hormone binding globulin (SHBG). The ‘benchmark set of steroids’ is a set of steroids with known affinities for human SHBG that has been widely used for validation in the development of different virtual screening methods. We have updated this data set by including additional steroidal SHBG ligands and by modifying the predicted binding orientations of several benchmark steroids in the SHBG binding site based on the use of an improved docking protocol and information from recent crystallographic data. The new steroid binding orientations and the expanded version of the benchmark set was then used to create new in silico models which were applied in virtual screening to identify high-affinity non-steroidal human SHBG ligands from a large chemical database. Anthropogenic compounds with the capacity to interact with the steroid-binding site of SHBG pose health risks to humans and other vertebrates including fish. We constructed a homology model of SHBG from zebrafish and applied virtual screening to identify ligands for zebrafish SHBG from a set of 80 000 existing commercial substances, many of which can be exposed to the aquatic environment. Six hits from this in silico screen were tested experimentally for zebrafish SHBG binding and three of them, hexestrol, 4-tert-octylcatechol, dihydrobenzo(a)pyren-7(8H)-one demonstrated micromolar binding affinity for the zebrafish SHBG. These findings demonstrate the feasibility of using virtual screening to identify anthropogenic compounds that may disrupt or highjack functionally important protein:ligand interactions. Studies applying this new computational toxicology method could increase the awareness of hazards posed by existing commercial chemicals at relatively low cost.

Tese (doutorado) - Universidade Federal de Santa Catarina, Centro de Ciências Físicas e Matemáticas, Programa de Pós-Graduação em Química, Florianópolis, 2009. / Made available in DSpace on 2012-10-24T22:16:21Z (GMT). No. of bitstreams: 1 267263.pdf: 2460883 bytes, checksum: de925adc7ffd8300f1031855c1454407 (MD5) / O presente trabalho descreve a síntese e atividade antinociceptiva de seis séries de chalconas, derivados imídicos e adutos de Diels-Alder, além da atividade antiparasitária de uma dessas séries. Todos os compostos foram testados em um modelo de nocicepção (os adutos foram avaliados, ainda, quanto à atividade antifúngica). A série inicial foi, também, avaliada quanto ao perfil antiprotozoário e apresentou excelentes resultados. Os adutos imídicos obtidos das reações de Diels-Alder mostraram fraca ação antinociceptiva, e razoável ação antifúngica. As chalconas e derivados apresentaram bom perfil analgésico, o que motivou a busca pela quantificação da relação estrutura-atividade. O primeiro método, para a quantificação da relação-estrutura atividade (QSAR), considerou os compostos pertencentes à primeira série de chalconas. Obtendo, para esta série, uma equação de QSAR com capacidade de predição de 72,2%,. Na segunda abordagem, foram comparadas as estruturas de seis séries de chalconas com relação aos resultados de atividade antinociceptiva para avaliação da QSAR. Essas séries foram divididas em dois grupos estruturais que possibilitaram a obtenção de mais duas equações de QSAR. Ambas apresentaram graus de predição muito bons, acima dos 70% (76,4% e 70,4%, respectivamente). Tais resultados revelam grande potencialidade de aplicação biológica dos compostos sintetizados.

Submitted by Pós graduação Farmácia (ppgfar@ufba.br) on 2017-05-25T19:31:51Z No. of bitstreams: 1 HUMBERTO.FREITAS2.pdf: 7925171 bytes, checksum: a90703d131530b19fcb2e6433ff99c07 (MD5) / Approved for entry into archive by Patricia Barroso (pbarroso@ufba.br) on 2017-05-29T17:25:44Z (GMT) No. of bitstreams: 1 HUMBERTO.FREITAS2.pdf: 7925171 bytes, checksum: a90703d131530b19fcb2e6433ff99c07 (MD5) / Made available in DSpace on 2017-05-29T17:25:44Z (GMT). No. of bitstreams: 1 HUMBERTO.FREITAS2.pdf: 7925171 bytes, checksum: a90703d131530b19fcb2e6433ff99c07 (MD5) / FAPESB / Apesar dos avanços no desenvolvimento de fármacos antifúngicos, observa-se um crescimento de cepas resistentes aos fármacos de escolha (derivados azólicos). Diante disso, torna-se imprescindível racionalizar as propriedades físico-químicas e estruturais que governam a atividade antifúngica de derivados azólicos, bem como investigar os mecanismos moleculares responsáveis pela resistência. Visando elucidar algumas dessas questões os objetivos desse trabalho são: 1) Determinar a atividade biológica de 33 derivados azólicos, frente á Cryptococcus gattii, a fim de utilizá-la como variável dependente na construção de modelos classificatórios (KNN e SIMCA) e quantitativos (HQSAR e QSAR 2D); 2) Investigar, do ponto de vista estrutural, mutações pontuais no gene ERG11 que conferem resistência a derivados azólicos. Os modelos classificatórios (KNN e SIMCA) mostram consistência interna razoável ao acertar a classificação de 60% dos compostos ativos e 75% dos compostos inativos do grupo teste. Adicionalmente, o modelo SIMCA sugere que a transferência de carga entre átomos distantes de 10 ligações (JGI10) é importante para diferenciar moléculas ativas das inativas contra C. gattii. Os modelos quantitativos baseados em fragmentos moleculares (hologramas QSAR) apresentam bom ajuste (r2=0,85), porém baixo poder preditivo (r2 pred=0,38), enquanto o melhor modelo de QSAR 2D baseado em descritores topológicos apresenta resultados estatísticos elevados (r2=0,95 e q2=0,86) com 3 PCs, além de bom poder preditivo (r2 pred=0,72). Este modelo indica ainda que o aumento da potência é influenciado pela menor distribuição de carga de uma molécula na distância de dois átomos (GGI1) e pela menor diferença de eletronegatividade dos átomos na distância topológica 1 e 2 (GATS1e e MATS2e). Os modelos comparativos de lanosterol 14α-desmetilase de C. gattii, nativa e com mutações pontuais (G484S, H488Y e K156R), mostram que alterações no posicionamento do grupo heme e redução do número de interações entre a enzima e o grupo prostético, desestabilizam a interação entre o ferro do grupo heme e o nitrogênio do anel imidazólico ou triazólico, e modificações na flexibilidade conformacional do alvo terapêutico, reduzindo assim o acesso do antifúngico ao sítio ativo da enzima, são os principais mecanismos pelos quais as mutações reduzem a afinidade dos derivados azólicos pelo alvo terapêutico. / Despite advances in the development of antifungal drugs, there is an upsurge of resistant strains against the drugs of choice (azole derivatives). Hence, it becomes essential to comprehend the physicochemical and structural properties that rule the antifungal activity of azole’s derivatives, as well as to investigate the molecular mechanisms responsible for fungal resistance. Aiming at elucidating some of these questions the objectives of this study are: 1) To determine the biological activity of 33 azole derivatives against Cryptococcus gattii, in order to use it as a dependent variable in the development of classification (KNN and SIMCA) and quantitative (HQSAR and 2D QSAR) models; 2) To investigate, from a structural stand-point of view, mutations in the ERG11 gene which render C. gattii resistant to azole’s derivatives. Classification models (KNN and SIMCA) show reasonable internal consistency, correctly classifying 60% of the active compounds and 75% of inactive compounds from the test group. Additionally, the SIMCA model suggests that the charge transfer between atoms 10 bonds apart (JGI10) is important to differentiate azole derivatives that are either active or inactive against C. gattii. The quantitative models based on molecular fragments (hologram QSAR) shows good fit (r2 = 0.85), but low predictive power (r2 pred = 0.38), whereas the best 2D QSAR model, built from topological descriptors, shows outstanding statistical results (r2 = 0.95 and q2 = 0.86) with 3 PCs, along with good predictive power (r2 pred = 0.72). This model also indicates that potency boost is influenced by reduced charge distribution between two atoms (GGI1) and by the lower electronegativity difference at the topological distance 1 and 2 (GATS1e and MATS2e). Comparative models of native and mutated (G484S, H488Y and K156R) lanosterol 14α- demethylases of C. gattii show that changes in the positioning of the heme group and reduced binding of the enzyme towards the prosthetic group, disrupt the interaction profile between the heme iron and the nitrogen of imidazole or triazole ring, as well as changes in the therapeutic target conformational flexibility, which reduces the access of the antifungal agente to the enzyme active site, are the main mechanisms by which affinity of azole derivatives is reduced.

Tese (doutorado) - Universidade Federal de Santa Catarina, Centro de Ciências Físicas e Matemáticas. Programa de Pós-Graduação em Química. / Made available in DSpace on 2012-10-20T15:08:00Z (GMT). No. of bitstreams: 1 198987.pdf: 5942141 bytes, checksum: 7bb584e2f7fedf5689044108457ed464 (MD5) / Neste estudo um novo descritor molecular - Índice Semi-Empírico Topológico (IET) - foi desenvolvido, a fim de estabelecer correlações quantitativas entre estrutura e propriedade (QSPR), para diferentes classes de compostos. Este Índice foi desenvolvido e otimizado para prever a retenção cromatográfica de alcenos ramificados, alcanos metil ramificados produzidos por insetos e álcoois saturados, em fases estacionárias de baixa polaridade. Foi avaliada, também, a habilidade de previsão do IET para a retenção cromatográfica de álcoois, aldeídos e cetonas em fases estacionárias mais polares. Os estudos preliminares aplicando o IET a diferentes propriedades/atividades apresentaram resultados promissores para a aplicação futura deste novo método. Para alcenos e álcoois foram obtidas correlações entre o IET e as propriedades (ponto de ebulição normal, refração molar, volume molar, calor de combustão, calor de vaporização molar e coeficiente de partição octanol/água), com valores de r > 0,94. As correlações quantitativas estrutura-atividade (QSAR) foram testadas para álcoois saturados, onde as atividades biológicas investigadas foram: atividade narcótica sobre larvas das cracas, toxicidade em aranhas e tomates e odor (r > 0,88). A qualidade dos resultados obtidos neste trabalho para a previsão de diferentes propriedades/atividades, empregando o IET como descritor molecular, pode ser considerada como uma importante etapa na direção de estudos futuros em QSAR/QSPR/QSRR.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Aedes aegypti is the main transmitter of vector-borne diseases such, dengue, chikungunya and zika. Hence no vaccine exists as well as drugs to reduce viremia, major strategy to prevent these diseases is controlling vector spreading. The main larvicides used are organophosphates and pyrethroids, however the indiscriminate use of these compounds gave rise Ae. aegypti resistant strains. A viable alternative to classic insecticides / larvicides is the phytochemical research of molecules, like terpenoids, component of essential oils of several plants that exhibits larvicidal activity. In this context, the aim of this study was synthetize and evaluate the activity of monoterpene (-)-borneol and twelve derivatives against Ae. aegypti larvae, using QSAR as strategy for discovery of new larvicidal agents. Reactions were carried out by esterification with acid chloride in basic medium, (-)-borneol was modificated in hydroxyl group, changing size and type of the side chain. Compounds synthetized were purified in column chromatography and characterized by NMR ¹³C, ¹H, MS and IR. LC50 was evaluated thought larvicidal assay, in each test, twenty third instar larvae were exposed to various concentrations of derivatives to 24h. From mortality data obtained, LC50 was determined though Probit analysis. The bornylchloroacetate derivative exhibited the best activity (21 ppm), but bornyl heptanoate showed no activity. The physicochemical properties of the derivatives were obtained by GAMESS® module Chem3D Ultra 7.0® software from most stable conformation of the molecule. Descriptor chosen for QSAR study was Log P, since best correlation obtained and especially to be a highly informative parameter to measure influence on the larvicidal activity. Equation was obtained by MiniTab16 ™ software by linear regression between derivatives Log P, and the activity expressed in log (1/LC50). QSAR equation without outliers exhibited quality indexes of r² = 0.944; F = 58.71; q² = 0.8442; Spress = 0.0827, indicating high predictability of the model. It was observed influence of lipophilicity on (-)-borneol derivatives larvicidal activity, suggesting that molecules with Log P value around 4.5 have optimized activity. This study may be used as a basis to guide research of new larvicides candidates. / O mosquito Aedes aegypti é o principal transmissor de doenças de origem viral, como a dengue, chikungunya e zika. Na ausência de recursos específicos como a falta de vacinas e medicamentos eficazes, a principal estratégia para o manejo destas infecções se dá pelo controle do vetor. Os principais agentes utilizados no controle químico são os organofosforados e piretroides, contudo o seu uso indiscriminado fez surgir populações do Ae. aegypti resistentes. Uma alternativa viável a estes inseticidas/larvicidas clássicos é a pesquisa de moléculas fitoquímicas, como as da classe dos terpenoides, presentes nos óleos essenciais de diversas espécies vegetais que já possuem atividade larvicida documentada na literatura. Neste contexto, o objetivo deste trabalho foi sintetizar e avaliar a atividade do monoterpeno (-)-borneol e seus doze derivados frente as larvas do Ae. aegypti, utilizando o QSAR como estratégia de descoberta de novos candidatos a agentes larvicidas. Através da reação de esterificação em meio básico, utilizando cloreto de ácido, o (-)-borneol foi modificado a partir de sua hidroxila alcoólica, variando no tamanho e tipo de cadeia lateral. Os compostos sintetizados foram purificados em coluna cromatográfica e caracterizados por RMN ¹³C e ¹H, EM, e IR. A CL50 foi avaliada através do ensaio larvicida, onde a cada teste 20 larvas em terceiro estádio são expostas por 24h a diferentes concentrações (em triplicata) de composto. A partir dos dados de mortalidade das larvas, a CL50 é obtida com IC 95% pela análise Probit. O derivado cloroacetato de bornila exibiu a maior atividade (21 ppm), já o heptanoato de bornila demonstrou-se inativo. As propriedades físico-químicas dos derivados foram obtidas pelo módulo GAMESS® do programa Chem3D Ultra 7.0® a partir da conformação mais estável da molécula. O descritor escolhido para estudo de QSAR foi o Log P, por apresentar a melhor correlação, e principalmente por ser um parâmetro altamente informativo quanto a sua influência na atividade larvicida. A equação foi calculada pelo software MiniTab16™ através da regressão linear entre o Log P dos derivados e a atividade larvicida expressa em Log (1/CL50). O QSAR obtido sem os compostos outliers apresentou índices de qualidade de r² = 0,944; F = 58,71; q²= 0,8442; SPRESS = 0,0827 indicando alta preditibilidade do modelo. Foi observada a influência da lipofilicidade na atividade larvicida dos derivados do (-)-borneol, sugerindo que moléculas com Log P de aproximadamente 4,5 tem sua atividade otimizada. Este trabalho poderá ser utilizado como base para direcionar o planejamento de novos candidatos a agentes larvicidas.

The emergence of pathogens resistant to available drug therapies is a pressing global health problem. Antimicrobial peptides (AMPs) may potentially form new therapeutics to counter these pathogens. AMPs are key components in the mammalian innate immune system and are responsible for both direct killing and immunomodulatory effects in host defense against pathogenic organisms. This thesis describes computational methods for the identification of novel natural and synthetic AMPs. A bioinformatic resource was constructed for classification and discovery of gene- coded AMPs, consisting of a database of clustered known AMPs and a set of hidden Markov models (HMMs). One set of 146 clusters was based on the mature peptide sequence, and one set of 40 clusters was based on propeptide sequence. The bovine genome was analyzed using the AMPer resources, and 27 of the 34 known bovine AMPs were identified with high confidence and up to 69 AMPs were predicted to be novel peptides. One novel cathelicidin AMP was experimentally verified as up-regulated in response to infection in bovine intestinal tissue. A chemoinformatic analysis was performed to model the antibacterial activity of short synthetic peptides. Using high-throughput screening data for the activities of over 1400 peptides of diverse sequence, quantitative structure-activity relation (QSAR) models were created using artificial neural networks and physical characteristics of the peptide that included three-dimensional atomic structure. The models were used to predict the activity of a set of approximately 100,000 peptide sequence variants. After ranking the predicted activity, the models were shown to be very accurate. When 200 peptides were synthesized and screened using four levels of expected activity, 94% of the top 50 peptides expected to have the highest level of activity were found to be highly active. Several promising candidates were synthesized with high quality and tested against several multi- antibiotic-resistant pathogens including clinical strains of Pseudomonas aeruginosa, Staphylococcus aureus, Enterococcus faecalis and Escherichia coli. These peptides were found to be highly active against these pathogens as determined by minimal inhibitory concentration; this serves as independent confirmation of the effectiveness of high-throughput screening and in silico analysis for identifying peptide antibiotic drug leads. / Medicine, Faculty of / Medicine, Department of / Experimental Medicine, Division of / Graduate

Virtual screening is a fast, low cost method to identify potential small molecule therapeutics from large chemical databases for the vast amount of target proteins emerging from the life sciences and bioinformatics. In this work, we applied several conventional and newly developed virtual screening approaches to identify novel non-steroidal ligands for the human and zebrafish sex hormone binding globulin (SHBG). The ‘benchmark set of steroids’ is a set of steroids with known affinities for human SHBG that has been widely used for validation in the development of different virtual screening methods. We have updated this data set by including additional steroidal SHBG ligands and by modifying the predicted binding orientations of several benchmark steroids in the SHBG binding site based on the use of an improved docking protocol and information from recent crystallographic data. The new steroid binding orientations and the expanded version of the benchmark set was then used to create new in silico models which were applied in virtual screening to identify high-affinity non-steroidal human SHBG ligands from a large chemical database. Anthropogenic compounds with the capacity to interact with the steroid-binding site of SHBG pose health risks to humans and other vertebrates including fish. We constructed a homology model of SHBG from zebrafish and applied virtual screening to identify ligands for zebrafish SHBG from a set of 80 000 existing commercial substances, many of which can be exposed to the aquatic environment. Six hits from this in silico screen were tested experimentally for zebrafish SHBG binding and three of them, hexestrol, 4-tert-octylcatechol, dihydrobenzo(a)pyren-7(8H)-one demonstrated micromolar binding affinity for the zebrafish SHBG. These findings demonstrate the feasibility of using virtual screening to identify anthropogenic compounds that may disrupt or highjack functionally important protein:ligand interactions. Studies applying this new computational toxicology method could increase the awareness of hazards posed by existing commercial chemicals at relatively low cost. / Science, Faculty of / Graduate

Orientadores: Yuji Takahata, Rogerio Custodio / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Quimica / Made available in DSpace on 2018-08-14T14:03:54Z (GMT). No. of bitstreams: 1 Oliveira_KesleyMoraesGodinhode_D.pdf: 5766531 bytes, checksum: 5d60e17b9a5062a054d9bb7de66054f4 (MD5) Previous issue date: 2009 / Resumo: O objeto de estudo desta tese é a forma mais severa e letal de Leishmaniose: a Leishmaniose Visceral, LV, cujo principal agente etiológico é a espécie Leishmania donovani. O objetivo deste trabalho é o desenvolvimento de modelos Quantitativos das Relações Estrutura-Atividade para duas séries de compostos com atividade antileishmaniose contra formas amastigotas de Leishmania donovani. A primeira série envolve vinte e um análogos de nucleosídeos pirazolo-pirimidínicos e a segunda série compreende oito antifúngicos. Para garantir a robustez dos modelos, além dos compostos que compõem as respectivas séries, foram selecionadas moléculas com conhecida atividade contra LV para compor um conjunto de teste e avaliar a capacidade de previsão dos respectivos modelos. Para a série dos nucleosídeos foram desenvolvidos modelos de regressão logística e árvore de classificação. Em ambas as abordagens os descritores Mor26v e o GAP(HOMO, HOMO-1) se mostraram relevantes para a explicação da atividade leishmanicida destes compostos. O modelo de regressão logística atingiu 90,5% para acurácia de classificação para o conjunto de trabalho e 58% para o conjunto de teste após a análise do domínio de aplicabilidade do modelo. O modelo para árvore de classificação alcançou 95% para acurácia de classificação para o conjunto de trabalho e 83% para o conjunto de teste. Para a série dos antifúngicos foi utilizado um modelo de regressão linear múltipla onde a energia eletrônica e a área da superfície polar se mostraram importantes para a atividade leishmanicida da série. Os valores previstos exibem 98% de correlação com os valores experimentais. Os valores encontrados para o conjunto de teste também estão de acordo com a literatura. Finalmente, novos compostos foram propostos para síntese e avaliação da atividade leishmanicida. / Abstract: The object of study of this thesis is the most severe and lethal form of leishmaniasis: visceral leishmaniasis, LV, whose main etiological agent is the species Leishmania donovani. The goal of this work is the development of Quantitative Structure-Activity Relationship models for two series of compounds presenting antileishmanial activity against amastigotes of Leishmania donovani. The first series includes twenty-one analogues of pyrazolo-pyrimidine nucleosides and the second series comprises eight antifungals. To ensure the robustness of the models, in addition to the compounds that make up their series, molecules with known activity against LV were selected to compose a testset and evaluate the predictive power of their models. For the series of nucleosides logistic regression models and tree classification were developed. In both approaches, the Mor26v and GAP(HOMO, HOMO-1) descriptors were relevant to explain the leishmanicidal activity of these compounds. The logistic regression model reached 90.5% for classification accuracy to the workingset and 58% for testset after analysis of the domain of applicability of the model. The classification tree model reached 95% for classification accuracy of the workingset and 86% for the testset. A multiple linear regression model was applied to the series of antifungal agents. The electron energy and polar surface area were important for the leishmanicidal activity of this series. The predicted values showed 98% of correlation with the experimental values. The values found for the testset are also in agreement with the literature. Finally, new compounds were proposed for synthesis and evaluation of leishmanicidal activity. / Doutorado / Físico-Química / Doutor em Ciências

Quantitative Structure-Activity Relationship (QSAR) has been applied extensively in predicting toxicity of Disinfection By-Products (DBPs) in drinking water. Among many toxicological properties, acute and chronic toxicities of DBPs have been widely used in health risk assessment of DBPs. These toxicities are correlated with molecular properties, which are usually correlated with molecular descriptors. The primary goals of this thesis are: 1) to investigate the effects of molecular descriptors (e.g., chlorine number) on molecular properties such as energy of the lowest unoccupied molecular orbital (ELUMO) via QSAR modelling and analysis; 2) to validate the models by using internal and external cross-validation techniques; 3) to quantify the model uncertainties through Taylor and Monte Carlo Simulation. One of the very important ways to predict molecular properties such as ELUMO is using QSAR analysis. In this study, number of chlorine (NCl) and number of carbon (NC) as well as energy of the highest occupied molecular orbital (EHOMO) are used as molecular descriptors. There are typically three approaches used in QSAR model development: 1) Linear or Multi-linear Regression (MLR); 2) Partial Least Squares (PLS); and 3) Principle Component Regression (PCR). In QSAR analysis, a very critical step is model validation after QSAR models are established and before applying them to toxicity prediction. The DBPs to be studied include five chemical classes: chlorinated alkanes, alkenes, and aromatics. In addition, validated QSARs are developed to describe the toxicity of selected groups (i.e., chloro-alkane and aromatic compounds with a nitro- or cyano group) of DBP chemicals to three types of organisms (e.g., Fish, T. pyriformis, and P.pyosphoreum) based on experimental toxicity data from the literature. The results show that: 1) QSAR models to predict molecular property built by MLR, PLS or PCR can be used either to select valid data points or to eliminate outliers; 2) The Leave-One-Out Cross-Validation procedure by itself is not enough to give a reliable representation of the predictive ability of the QSAR models, however, Leave-Many-Out/K-fold cross-validation and external validation can be applied together to achieve more reliable results; 3) ELUMO are shown to correlate highly with the NCl for several classes of DBPs; and 4) According to uncertainty analysis using Taylor method, the uncertainty of QSAR models is contributed mostly from NCl for all DBP classes.