Identification and validation of Perioperative anesthesia -associated dignostic and druggable frameworks for Deep vein thrombosis patients and its associations with Kawasaki disease: insights from machine learning and multi-omics

Methods

By combination of 3 DVE patient bulk profiles(GSE118259, GSE19151 and GSE48000) with perioperative anesthesia-related drug target(PARDTGs) and integrative bioinformatic analysis(Limma and machine learning framework), we identified hub differentially expressed PARDTGs for DVE patients. The dignostic performance and corresponding molecular and immune features were also estimated among DVE patients. Furthermore, natural compounds for alleviation of DVE targeting hub PARDTGs were also enriched by CTD database and validated by molecular docking.

In addition, differentially expressed PARDTGs can guide the risk stratification for DVE patients via consensus clustering. Besides, KD peripheral blood single-cell profile(GSE254657) was utilized, and hub differentially expressed PARDTGs functional performance in investigation of association between DVT and patients were estimated in spatial and temporal manners.