CoBDock is a novel blind docking method that integrates docking and cavity detection results using machine learning to improve binding site identification and pose prediction accuracy. It uses four molecular docking algorithms (AutoDock Vina, GalaxyDock3, ZDOCK, and PLANTS) and two cavity detection tools (P2Rank and Fpocket) to identify binding sites and predict ligand poses. The method processes 3D structural data into grids and uses a machine learning model to rank voxels and select the most likely binding site. CoBDock has been tested on several benchmark datasets, including PDBBind 2020, ADS, MTi, DUD-E, and CASF-2016, and has shown superior performance compared to other state-of-the-art cavity detection tools and blind docking methods. The method uses a consensus approach to combine results from multiple docking and cavity detection tools, enhancing the accuracy of binding site identification and pose prediction. CoBDock also provides an automated end-to-end pipeline for docking multiple ligands to multiple targets, improving its practicality. The method has been shown to outperform other methods in terms of binding site identification and pose prediction accuracy across various benchmarks.CoBDock is a novel blind docking method that integrates docking and cavity detection results using machine learning to improve binding site identification and pose prediction accuracy. It uses four molecular docking algorithms (AutoDock Vina, GalaxyDock3, ZDOCK, and PLANTS) and two cavity detection tools (P2Rank and Fpocket) to identify binding sites and predict ligand poses. The method processes 3D structural data into grids and uses a machine learning model to rank voxels and select the most likely binding site. CoBDock has been tested on several benchmark datasets, including PDBBind 2020, ADS, MTi, DUD-E, and CASF-2016, and has shown superior performance compared to other state-of-the-art cavity detection tools and blind docking methods. The method uses a consensus approach to combine results from multiple docking and cavity detection tools, enhancing the accuracy of binding site identification and pose prediction. CoBDock also provides an automated end-to-end pipeline for docking multiple ligands to multiple targets, improving its practicality. The method has been shown to outperform other methods in terms of binding site identification and pose prediction accuracy across various benchmarks.