KICDB: A Causality-Oriented Multi-Omics Database for Kinase Inhibitor-Induced Cardiotoxicity

BackgroundKinase inhibitors (KIs) are mainstays of targeted cancer therapy, but their clinical utility is frequently limited by cardiotoxicity. A systematic resource to explore the underlying causal mechanisms is urgently needed. MethodsWe present the KICDB (Kinase Inhibitor Cardiotoxicity Database), a comprehensive and interactive web server. KICDB is built upon a framework integrating large-scale transcriptomics meta-analysis with causal inference. ResultsThis database centralizes the findings from a comprehensive meta-analysis of 26 kinase inhibitors (KIs) across 7 studies (n=5291) identified 8,907 significant gene expression changes in human cardiomyocytes. To establish causality, we performed a two-pronged Mendelian randomization (MR) analysis testing hundreds of downstream genes and a panel of 43 key kinase proteins against 46 cardiovascular outcomes. This large-scale analysis revealed 26 significant causal associations, implicating novel molecular mediators in KI-induced cardiotoxicity. ConclusionsKICDB serves as a valuable and accessible platform for the cardio-oncology community. By integrating transcriptomic signatures with causal inference data, the database empowers researchers to formulate mechanistic hypotheses, accelerate biomarker validation, and guide the design of future cardioprotective strategies. URL: https://zhang-lab-database.shinyapps.io/KICDB/ Key PointsO_LIWe developed KICDB, a comprehensive and publicly accessible web server, to systematically investigate the causal mechanisms of KI-induced cardiotoxicity. C_LIO_LIKICDB integrates a large-scale meta-analysis of transcriptomic data from 26 KIs with a robust Mendelian randomization (MR) framework to move beyond correlation and infer causality. C_LIO_LIThe analysis identified 8,907 significant gene expression changes and 26 significant causal associations between KI-associated genes and 46 cardiovascular outcomes. C_LI