Supporting data for: In Silico Identification of Novel Phytochemicals Targeting Canine Melanoma on Homology-Modeled Proteins

Canine melanoma lacks experimental structures for key drug targets PI3Kalpha and MEK1. We generated homology models, validated with SAVES, and screened 13 ligands using Glide extra precision docking. Binding affinities were refined with MM-GBSA calculations. Molecular dynamics simulations assessed RMSD and RMSF. The homology models exhibited good stereochemical quality. Luteolin showed balanced dual-target binding. Molecular dynamics revealed RMSD values within 2-3 Angstrom and RMSF below 2 Angstrom at binding residues. ADME profiling indicated favorable drug-like properties with no Lipinski violations. Luteolin represents a potential dual inhibitor with balanced binding affinity and stable complex dynamics in silico, warranting further experimental evaluation.犬黑色素瘤的关键药物靶点PI3Kα（PI3Kalpha）与MEK1暂无相关实验结构。本研究构建了同源模型，并通过SAVES工具完成验证；随后采用Glide超精确对接对13种配体开展虚拟筛选，结合亲和力通过MM-GBSA计算完成精细化优化。通过分子动力学模拟对均方根偏差（RMSD）与均方根波动（RMSF）进行评估，结果显示所构建的同源模型具备良好的立体化学质量。其中木犀草素（Luteolin）展现出均衡的双靶点结合特性，分子动力学模拟表明其结合位点残基处的RMSD值处于2~3埃范围内，RMSF值低于2埃。ADME（吸收-分布-代谢-排泄）性质分析显示，该化合物具有优良的类药性质，未违反Lipinski法则。综上，木犀草素是一种兼具均衡结合亲和力与稳定复合物动力学特性的潜在双靶点抑制剂，值得开展进一步的实验验证。