Scholarlink

<p>Psoriatic arthritis (PsA) is an immune-mediated inflammatory arthropathy occurring as a consequence of psoriasis that presents with chronic joint pain, swelling, stiffness and progressive structural damage along with enthesitis, dactylitis and reduced mobility. Management strategies to treat joint pain range from NSAIDs, corticosteroids as well as conventional and biologic DMARDs.Among these, phosphodiesterase 4 (PDE4) inhibitors represent a key oral non-biologic treatment option although approved agents such as Apremilast are limited by adverse effects. In this study, bioactivity data for PDE4 family inhibitors were retrieved from ChEMBL and processed into molecular descriptors and fingerprints for quantitative structure–activity relationship (QSAR) modelling. Feature refinement using correlation filtering and Random Forest based importance ranking resulted in a reduced biologically meaningful feature set. The final predictive model demonstrated stable performance supported by internal validation and cross validation. Chemical space was further explored using KMeans clustering enabling identification of more coherent structure activity relationships. Scaffolds extracted from high-performing clusters were subsequently screened against the COCONUT natural product database yielding 629 structurally relevant hits. Prioritized compounds were evaluated using molecular docking against PDE4B (PDB ID: 1XMU). Two candidates, hyrtioseragamine B (CNP0154829.0) and amodiaquine (CNP0349391.0), exhibited favourable binding affinities and interaction patterns comparable to the reference inhibitor apremilast. Subsequent in silico ADME profiling suggested complementary advantages: amodiaquine displayed favourable drug-likeness, while hyrtioseragamine B, despite its structural complexity, demonstrated promising binding characteristics and may represent a novel chemotype for PDE4B inhibition.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning integrated QSAR and molecular docking for identifying natural product derived PDE4B inhibitors in psoriatic arthritis

  • 期刊论文

  • 发表日期：

    2026年7月14日

查看全文
  • S Raghavendra,
  • M Pooja,
  • Rajat Kulshreshtha,
  • V. G. Shanmuga Priya

摘要

Psoriatic arthritis (PsA) is an immune-mediated inflammatory arthropathy occurring as a consequence of psoriasis that presents with chronic joint pain, swelling, stiffness and progressive structural damage along with enthesitis, dactylitis and reduced mobility. Management strategies to treat joint pain range from NSAIDs, corticosteroids as well as conventional and biologic DMARDs.Among these, phosphodiesterase 4 (PDE4) inhibitors represent a key oral non-biologic treatment option although approved agents such as Apremilast are limited by adverse effects. In this study, bioactivity data for PDE4 family inhibitors were retrieved from ChEMBL and processed into molecular descriptors and fingerprints for quantitative structure–activity relationship (QSAR) modelling. Feature refinement using correlation filtering and Random Forest based importance ranking resulted in a reduced biologically meaningful feature set. The final predictive model demonstrated stable performance supported by internal validation and cross validation. Chemical space was further explored using KMeans clustering enabling identification of more coherent structure activity relationships. Scaffolds extracted from high-performing clusters were subsequently screened against the COCONUT natural product database yielding 629 structurally relevant hits. Prioritized compounds were evaluated using molecular docking against PDE4B (PDB ID: 1XMU). Two candidates, hyrtioseragamine B (CNP0154829.0) and amodiaquine (CNP0349391.0), exhibited favourable binding affinities and interaction patterns comparable to the reference inhibitor apremilast. Subsequent in silico ADME profiling suggested complementary advantages: amodiaquine displayed favourable drug-likeness, while hyrtioseragamine B, despite its structural complexity, demonstrated promising binding characteristics and may represent a novel chemotype for PDE4B inhibition.