In silico AI-guided de novo design of S1PR1-targeting ligands from a siponimod scaffold for angiogenesis and fat-cartilage graft survival

Fat cartilage graft survival is frequently limited by insufficient early vascularization, resulting in ischemia and partial tissue loss. Sphingosine-1-phosphate receptor 1 (S1PR1) is a key regulator of endothelial barrier integrity, vascular maturation, and controlled angiogenic signaling. Modulation of S1PR1 has been associated with stabilization of vascular networks and regulation of endothelial sprouting, making it a potential therapeutic target for structured angiogenesis. Targeted modulation of endogenous vascular regulatory pathways may provide a strategy to enhance structured angiogenesis while maintaining vascular stability. In this study, an integrated AI-guided and structure-based computational framework was applied to design and prioritize novel sphingosine-1-phosphate receptor 1 (S1PR1)-binding ligands derived from a siponimod scaffold. A convolutional neural network-based generative model produced a library of candidate molecules, which were evaluated using dual-platform molecular docking, in silico ADMET profiling, quantum chemical analysis, and long-timescale molecular dynamics simulation. Among the generated compounds, AI Derivative 1 consistently achieved a top-ranked docking score (~ −10.2 kcal/mol) comparable to the reference ligand siponimod. Pharmacokinetic predictions indicated compliance with Lipinski criteria and favorable predicted gastrointestinal absorption. Density functional theory analysis showed electronic properties comparable to siponimod, supporting physicochemical consistency. Comparative 1 μs molecular dynamics simulations indicated persistent accommodation of both ligands under the modeled conditions. MM-GBSA and MM-PBSA produced favorable endpoint binding-energy estimates for AI Derivative 1, although siponimod was ranked as energetically more favorable by both models. However, docking scores and endpoint free energy calculations were used exclusively for comparative ranking and do not establish functional agonism or angiogenic efficacy. Collectively, these results identify AI Derivative 1 as a computationally prioritized S1PR1-binding candidate for future experimental evaluation. Biological activity, receptor signaling behavior, and therapeutic relevance remain to be established through in vitro and in vivo studies.