Prediction of the Radical Scavenging Activities of Some Antioxidant from Their Molecular Structure

Fatemi, M.; Rostami, Elham Gholami

doi:10.1021/ie4001426

Cited by 9 publications

(7 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous papers have presented promising predictions for antioxidative properties of small molecules 18 , proteins [19][20][21] and peptides [22][23][24] using different machine learning algorithms (e.g. Multiple Linear Regression, Support Vector Machines and Random Forest).…”

mentioning

confidence: 99%

AnOxPePred: using deep learning for the prediction of antioxidative properties of peptides

Olsen

Yeşiltaş

Marin

et al. 2020

Sci Rep

102

View full text Add to dashboard Cite

Dietary antioxidants are an important preservative in food and have been suggested to help in disease prevention. With consumer demands for less synthetic and safer additives in food products, the food industry is searching for antioxidants that can be marketed as natural. Peptides derived from natural proteins show promise, as they are generally regarded as safe and potentially contain other beneficial bioactivities. Antioxidative peptides are usually obtained by testing various peptides derived from hydrolysis of proteins by a selection of proteases. This slow and cumbersome trial-and-error approach to identify antioxidative peptides has increased interest in developing computational approaches for prediction of antioxidant activity and thereby reduce laboratory work. A few antioxidant predictors exist, however, no tool predicting the antioxidative properties of peptides is, to the best of our knowledge, currently available as a web-server. We here present the AnOxPePred tool and web-server (http://services.bioinformatics.dtu.dk/service.php?AnOxPePred-1.0) that uses deep learning to predict the antioxidant properties of peptides. Our model was trained on a curated dataset consisting of experimentally-tested antioxidant and non-antioxidant peptides. For a variety of metrics our method displays a prediction performance better than a k-NN sequence identity-based approach. Furthermore, the developed tool will be a good benchmark for future predictors of antioxidant peptides.

show abstract

mentioning

confidence: 99%

AnOxPePred: using deep learning for the prediction of antioxidative properties of peptides

Olsen

Yeşiltaş

Marin

et al. 2020

Sci Rep

102

View full text Add to dashboard Cite

show abstract

“…Figures (a) and (b) show the plots of predicted versus experimental values of log K for models 1 and 2, respectively, and indicate that there is a linear relationship between the adsorption coefficients of small organic molecules onto MWCNTs and the optimal descriptors of CORAL. To ensure that there is no chance correlation between descriptor matrix and vector of log K , Y ‐scrambling test was carried out . This test was formulated by Mitra and Roy and was corrected by Todeschini as follows:

{}^{C}R_{P}^{2} = R \sqrt{(R^{2} - R_{r}^{2})}

R_{r}^{2}

is average of R 2 after many times scrambling of log K vectors and modeling.…”

Section: Resultsmentioning

confidence: 99%

A Theoretical Approach to Model and Predict the Adsorption Coefficients of Some Small Aromatic Molecules on Carbon Nanotube

Heidari

Fatemi

2017

J Chinese Chemical Soc

Self Cite

View full text Add to dashboard Cite

In this study, for the first time, the CORAL software is used to model the quantitative structureproperty relationship (QSPR) of adsorption coefficients of chemicals on carbon nanotubes. The new method, which includes three-dimensional response surfaces for all subsets, is used to optimize Monte Carlo parameters in the modeling procedure. By using the 3D response surfaces, we obtained a satisfactory value for R 2 test , at the same time keeping the statistical quality of the subtraining and validation sets. Two predictive models were developed by using hydrogen-filled graph-based descriptors (model 1) and hybrid descriptors (model 2) with squared correlation coefficient (R 2 ) values of 0.985 and 0.940, respectively. The prediction power of these models was evaluated on a six-member test set, leading to R 2 values of 0.962 and 0.941 for model 1 and model 2, respectively. Moreover, the leave-one-out cross validation test was used to investigate the robustness of models, which lead to Q 2 values of 0.983 for model 1 and 0.932 for model 2.

show abstract

“…Li et al have used MLP-ANN model to predict the antioxidant activity of polysaccharides in DPPH test and used sensitivity analysis to interpret the effect of the input variables on the target values 23 . Petar et al and Fatemi et al have used ANN and MLP-ANN QSAR models to evaluate the contribution of the quantum mechanical molecular descriptors to the Trolox-equivalent antioxidant capacity (TEAC) in an optimized ANN model 19 , 24 . Although the prediction accuracy of ANN is higher than MLR, most of the current ANN methods used to predict antioxidant activity are more like a black box that has overfitting risk and lead to unreliable predictions.…”

Section: Introductionmentioning

confidence: 99%

Support vector regression-based QSAR models for prediction of antioxidant activity of phenolic compounds

Shi

2021

Sci Rep

View full text Add to dashboard Cite

The Support vector regression (SVR) was used to investigate quantitative structure–activity relationships (QSAR) of 75 phenolic compounds with Trolox-equivalent antioxidant capacity (TEAC). Geometric structures were optimized at the EF level of the MOPAC software program. Using Pearson correlation coefficient analysis, four molecular descriptors [n(OH), Cosmo Area (CA), Core-Core Repulsion (CCR) and Final Heat of Formation (FHF)] were selected as independent variables. The QSAR model was developed from the training set consisting of 57 compounds and then used the leave-one-out cross-validation (LOOCV) correlation coefficient to evaluate the prediction ability of the QSAR model. Used Artificial neural network (ANN) and multiple linear regression (MLR) for comparing. The RMSE (root mean square error) values of LOOCV in SVR, ANN and MLR models were 0.44, 0.46 and 0.54. The RMSE values of prediction of external 18 compounds were 0.41, 0.39 and 0.54 for SVR, ANN and MLR models, respectively. The obtained result indicated that the SVR models exhibited excellent predicting performance and competent for predicting the TEAC of phenolic compounds.

show abstract

Prediction of the Radical Scavenging Activities of Some Antioxidant from Their Molecular Structure

Cited by 9 publications

References 37 publications

AnOxPePred: using deep learning for the prediction of antioxidative properties of peptides

AnOxPePred: using deep learning for the prediction of antioxidative properties of peptides

A Theoretical Approach to Model and Predict the Adsorption Coefficients of Some Small Aromatic Molecules on Carbon Nanotube

Support vector regression-based QSAR models for prediction of antioxidant activity of phenolic compounds

Contact Info

Product

Resources

About