Integrated in-silico approach for identification of TNF-α inhibitors from an in-house library using 3D-QSAR, Molecular Dynamics, MM/GBSA, and WaterSwap analysis

• Developed robust 3D-QSAR and Activity Atlas models for TNF-α inhibitory activity. • Analysed SAR focusing on electrostatic, hydrophobic, and activity shape analysis. • Identified 80 molecules as TNF- α inhibitors with pIC 50 values ≥ 5.5. • Molecular docking and Dynamics suggested five most potent compounds. • Study offers valuable insights into SAR for development of potential TNF-α inhibitor. TNF-α is a pro-inflammatory cytokine involved in the progression of many complex disorders. The availability of orally effective small-molecule inhibitors remains limited. In this study, a field-point based 3D-QSAR model integrated with Activity Atlas was developed and validated using a dataset of 83 compounds (r² = 0.986, q² = 0.616, external r² = 0.725), providing mechanistic insights into the electrostatic, hydrophobic, and steric determinants of TNF-α inhibition. The model was applied to an in-house library of 216 structurally diverse compounds, leading to the identification of 80 potential inhibitors (pIC₅₀ ≥ 5.5). These compounds were prioritized using molecular docking, electrostatic complementarity, and MM/GBSA analysis, resulting in the identification of DDSL121 and DDSL133 as top candidates. Triplicate 200 ns molecular dynamics simulations supported the stability of the protein–ligand complexes, as indicated by consistent RMSD, RMSF, and radius of gyration profiles. WaterSwap analysis further supported the predicted binding affinities and revealed favorable interaction energetics. ADMET profiling indicated that DDSL121 exhibits favorable pharmacokinetic properties with lower predicted toxicity. While DDSL133 demonstrated more uniform binding stability across the TNF-α trimer, DDSL121 emerged as the most promising TNF-α inhibitor due to its stronger binding affinity and superior ADMET profile compared to the reference inhibitor SPD-304. Overall, this study integrates 3D-QSAR, Activity Atlas, MM/GBSA, molecular dynamics, and WaterSwap within a unified computational framework, providing mechanistic and structure–activity relationship insights that facilitate the rational design of potent and orally bioavailable TNF-α inhibitors.