Close attention has been paid to estrogen compounds because these chemicals may pose a serious threat to the health of humans and wildlife. Estrogen receptor (ER) exists as two subtypes, ERα and ERβ. The difference in amino acids sequence of the binding sites of ERα and ERβ might lead to a result that some synthetic estrogens and naturally occurring steroidal ligands have different relative affinities and binding modes for ERα and ERβ. In this investigation, comparative molecular similarity indices analysis (CoMSIA) was performed on 50 estrogen compounds binding ERβ to find out the structural relationship with the activities. We also compared two alignment schemes employed in CoMSIA analy-sis, namely, atom-fit and receptor-based alignment, with respect to the predictive capability of their respective models for structurally diverse data sets. The model with the significant correlation and the best predictive power (R2=0.961, qL 2OO=0.671, RP 2red=0.722) was achieved. The CoMSIA and docking results revealed the structural features related to an activity and provided an insight into molecular mechanisms of estrogenic activities for estrogen compounds.
YANG XuShuWANG XiaoDongLUO SiJI LiQIN LiangLI RongSUN ChengWANG LianSheng
Estrogen compounds are suspected of disrupting endocrine functions by mimicking natural hormones, and such compounds may pose a serious threat to the health of humans and wildlife. Close attention has been paid to the prediction and molecular mechanisms of estrogen activity for estrogen compounds. In this article, estrogen receptor α subtype (ERα)–based comparative molecular similarity indices analysis (COMSIA) was performed on 44 estrogen compounds with structural diversity to find out the structural relationship with the activity and to predict the activity. The model with the significant correlation and the best predictive power (R2 = 0.965, Q2LOO = 0.599, R2pred = 0.825) was achieved. The COMSIA and docking results revealed the structural features for estrogen activity and key amino acid residues in binding pocket, and provided an insight into the interaction between the ligands and these amino acid residues.
YANG XuShuWANG XiaoDongJI LiLI RongSUN ChengWANG LianSheng
Since the complexity and structural diversity of man-made compounds are considered, quantitative structure-activity relationships (QSARs)-based fast screening approaches are urgently needed for the assessment of the potential risk of endocrine disrupting chemicals (EDCs). The artificial neural net-works (ANN) are capable of recognizing highly nonlinear relationships, so it will have a bright applica-tion prospect in building high-quality QSAR models. As a popular supervised training algorithm in ANN, back-propagation (BP) converges slowly and immerses in vibration frequently. In this paper, a research strategy that BP neural network was improved by conjugate gradient (CG) algorithm with a variable selection method based on genetic algorithm was applied to investigate the QSAR of EDCs. This re-sulted in a robust and highly predictive ANN model with R2 of 0.845 for the training set, q2pred of 0.81 and root-mean-square error (RMSE) of 0.688 for the test set. The result shows that our method can provide a feasible and practical tool for the rapid screening of the estrogen activity of organic compounds.
JI LiWANG XiaoDongYANG XuShuLIU ShuShenWANG LianSheng
