Respiratory syncytial virus (RSV) is a leading cause of severe lower respiratory tract infections in infants, the elderly, and immunocompromised individuals worldwide. The pathogenic mechanism of RSV is closely linked to the membrane fusion process mediated by its fusion glycoprotein (F protein), which has consequently emerged as a critical target for developing anti-RSV therapeutics. At present, there is a lack of specific clinical treatments for RSV, and traditional drug discovery approaches are often time-consuming and expensive. In this context, quantitative structure–activity relationship (QSAR)–assisted drug design offers notable advantages. In this study, we collected a dataset consisting of 156 benzimidazole derivatives against F protein from publicly available sources. Transferable, reproducible, and interpretable 2D-QSAR inhibitory activity and cytotoxicity prediction models were constructed using Genetic Algorithm (GA) and Multiple Linear Regression (MLR). Following rigorous statistical validation, the best inhibitory activity model achieved R2 = 0.8740, <InlineEquation ID="IEq1"> <EquationSource Format="TEX">${Q}_{\text{Loo}}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msubsup> <mi>Q</mi> <mrow> <mtext>Loo</mtext> </mrow> <mn>2</mn> </msubsup> </math></EquationSource> </InlineEquation> = 0.8272, <InlineEquation ID="IEq2"> <EquationSource Format="TEX">${R}_{\text{test}}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msubsup> <mi>R</mi> <mrow> <mtext>test</mtext> </mrow> <mn>2</mn> </msubsup> </math></EquationSource> </InlineEquation> = 0.8273, <InlineEquation ID="IEq3"> <EquationSource Format="TEX">${Q}_{\text{Fn}}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msubsup> <mi>Q</mi> <mrow> <mtext>Fn</mtext> </mrow> <mn>2</mn> </msubsup> </math></EquationSource> </InlineEquation> = 0.8033–0.8492, CCCtest = 0.8782, MAEtest = 0.3014; the best cytotoxicity model was of R2 = 0.7573, <InlineEquation ID="IEq4"> <EquationSource Format="TEX">${Q}_{\text{Loo}}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msubsup> <mi>Q</mi> <mrow> <mtext>Loo</mtext> </mrow> <mn>2</mn> </msubsup> </math></EquationSource> </InlineEquation> = 0.6926, <InlineEquation ID="IEq5"> <EquationSource Format="TEX">${R}_{\text{test}}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msubsup> <mi>R</mi> <mrow> <mtext>test</mtext> </mrow> <mn>2</mn> </msubsup> </math></EquationSource> </InlineEquation> = 0.7707, <InlineEquation ID="IEq6"> <EquationSource Format="TEX">${Q}_{\text{Fn}}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msubsup> <mi>Q</mi> <mrow> <mtext>Fn</mtext> </mrow> <mn>2</mn> </msubsup> </math></EquationSource> </InlineEquation> = 0.7298–0.8656, CCCtest = 0.8639, MAEtest = 0.1342. The optimal inhibitory activity model was used to perform virtual screening on 912 benzimidazole derivatives retrieved from the PubChem, and identified 234 derivatives with better inhibitory activity than the reference JNJ-53718678. Among these, 152 derivatives were found to possess better docking binding energies than JNJ-53718678. Furthermore, we used the optimal toxicity model to assess their cytotoxicity, and identified 23 derivatives with predicted cytotoxicity lower than that of JNJ-53718678. Finally, through drug-likeness evaluation, ADMET analysis and molecular dynamics simulation, we obtained eight potential RSV inhibitors with higher inhibitory activity, lower cytotoxicity, and better pharmacokinetic properties compared to JNJ-53718678. Graphical abstract

Discovery of potential RSV fusion protein inhibitors from benzimidazole derivatives using QSAR, molecular docking, and ADMET evaluation methods

期刊论文
发表日期：

2025年9月25日

查看全文

Yini Xie,
Runqing Jia,
Tengjiao Fan,
Shuo Chen,
Ting Ren,
Na Zhang,
Lijiao Zhao,
Rugang Zhong,
Guohui Sun

摘要

Respiratory syncytial virus (RSV) is a leading cause of severe lower respiratory tract infections in infants, the elderly, and immunocompromised individuals worldwide. The pathogenic mechanism of RSV is closely linked to the membrane fusion process mediated by its fusion glycoprotein (F protein), which has consequently emerged as a critical target for developing anti-RSV therapeutics. At present, there is a lack of specific clinical treatments for RSV, and traditional drug discovery approaches are often time-consuming and expensive. In this context, quantitative structure–activity relationship (QSAR)–assisted drug design offers notable advantages. In this study, we collected a dataset consisting of 156 benzimidazole derivatives against F protein from publicly available sources. Transferable, reproducible, and interpretable 2D-QSAR inhibitory activity and cytotoxicity prediction models were constructed using Genetic Algorithm (GA) and Multiple Linear Regression (MLR). Following rigorous statistical validation, the best inhibitory activity model achieved R² = 0.8740, ${Q}_{\text{Loo}}^{2}$ $Q_{Loo}^{2}$ = 0.8272, ${R}_{\text{test}}^{2}$ $R_{test}^{2}$ = 0.8273, ${Q}_{\text{Fn}}^{2}$ $Q_{Fn}^{2}$ = 0.8033–0.8492, CCC_test = 0.8782, MAE_test = 0.3014; the best cytotoxicity model was of R² = 0.7573, ${Q}_{\text{Loo}}^{2}$ $Q_{Loo}^{2}$ = 0.6926, ${R}_{\text{test}}^{2}$ $R_{test}^{2}$ = 0.7707, ${Q}_{\text{Fn}}^{2}$ $Q_{Fn}^{2}$ = 0.7298–0.8656, CCC_test = 0.8639, MAE_test = 0.1342. The optimal inhibitory activity model was used to perform virtual screening on 912 benzimidazole derivatives retrieved from the PubChem, and identified 234 derivatives with better inhibitory activity than the reference JNJ-53718678. Among these, 152 derivatives were found to possess better docking binding energies than JNJ-53718678. Furthermore, we used the optimal toxicity model to assess their cytotoxicity, and identified 23 derivatives with predicted cytotoxicity lower than that of JNJ-53718678. Finally, through drug-likeness evaluation, ADMET analysis and molecular dynamics simulation, we obtained eight potential RSV inhibitors with higher inhibitory activity, lower cytotoxicity, and better pharmacokinetic properties compared to JNJ-53718678.

Graphical abstract