In-Silico Design, QSAR Modeling, Molecular Docking, MD Simulation, Pharmacokinetic and ADMET Studies of 3-Cyanopyridine Derivatives as Potential Anti-Breast Cancer Agents

Abstract

Background Breast cancer remains a major global health challenge, creating a need for new compounds with improved antiproliferative activity and drug-like properties. This study investigated 3-cyanopyridine derivatives with reported activity against the MCF-7 breast cancer cell line and explored their potential as PIM-1-targeting anticancer candidates. Methods A series of 3-cyanopyridine derivatives with reported experimental IC 50 values was subjected to QSAR modelling using PaDEL descriptors, Kennard–Stone dataset division, genetic function approximation (GFA) and multiple linear regression. Model performance was assessed using R², adjusted R², Q² CV , external R² pred , multicollinearity analysis and applicability-domain analysis. Y-scrambling was used to further assess the robustness of the QSAR model against chance correlation by randomly permuting the response variable (pIC₅₀) while retaining the original descriptor matrix and repeating model construction for 10 iterations. Five new analogues (A1–A5) were designed from the QSAR model. Their predicted interactions with PIM-1 were evaluated using Glide XP molecular docking, followed by 500-ns molecular dynamics simulations for A1 and A5 and in-silico pharmacokinetic/ADMET prediction. Results Four QSAR models were developed, and the first model was selected due to its robust internal and external statistical performances as follows: R² = 0.96165, R 2 adj = 0.952062, Q² CV = 0.934224 and R² pred = 0.749. The mean R² of the scrambled models was 0.2336, the mean Q² was − 0.3655, and the calculated cR²p was 0.8874. The designed analogues had predicted pIC₅₀ values of 6.045–7.030, with A1 showing the highest predicted activity (7.030). Applicability-domain assessment of the designed analogues identified A1 and A5 as belonging to the chemical space represented by the model. Redocking of VRV gave a heavy-atom RMSD of 0.86 Å. A5 showed the most favourable predicted Glide XP docking score (− 6.924 kcal mol⁻¹), compared with − 6.790 kcal mol⁻¹ for VRV and − 6.811 kcal mol⁻¹ for the refence compound AZD1208. The 500-ns simulations indicated stable protein–ligand behaviour for A1 and A5 under the simulated conditions. Predicted pharmacokinetic/ADMET profiles further supported A1 and A5 for further investigation as anti-breast cancer agents. Conclusions The integrated QSAR, docking, molecular dynamics and ADMET analyses identify A1 and A5 as computationally prioritised 3-cyanopyridine candidates for experimental synthesis, PIM-1 inhibition testing and MCF-7 antiproliferative evaluation.

Abstract as supplied by the indexing source. © the authors / publisher.