AI Integration of Multi-omics Data for Drug Target Prediction in Glioblastoma

Glioblastoma (GBM) is an aggressive primary brain tumor with marked molecular and cellular heterogeneity, and current treatment provides limited long-term disease control. Single-target therapies have frequently failed because tumor subpopulations can differ in signaling dependencies, adapt under treatment pressure, and interact with a complex microenvironment. This review examines how artificial intelligence (AI) can integrate multi-omics data to improve drug-target prediction in GBM. It summarizes the pathological and molecular basis of GBM, evaluates current therapeutic targets and their limitations, and compares genomic, transcriptomic, single-cell, spatial, proteomic, metabolomic, and imaging data used for target discovery. Machine-learning, deep-learning, and network-based approaches are discussed with examples of multimodal risk stratification and candidate-target identification. The review also evaluates CRISPR screening as a functional validation step that can distinguish statistical associations from genuine tumor dependencies. Overall, AI-integrated multi-omics can provide a more context-specific approach to target prioritization by linking candidate genes to patient subgroups, cell states, and pathway activity. However, small cohorts, missing data, batch effects, limited model generalizability, and insufficient experimental validation remain major barriers. Future studies should combine larger multi-centre datasets with longitudinal and spatial profiling, interpretable models, functional perturbation, and prospective validation to support clinically translatable precision therapy in GBM.