The intersection of radiogenomics and personalized radiotherapy has gained significant attention, particularly in understanding the long-term risks faced by childhood cancer survivors. A study by Gibson et al. analyzed data from 11,220 survivors, revealing that cancer-specific polygenic risk scores (PRSs) can predict the likelihood of developing subsequent cancers such as basal cell carcinoma and melanoma, with odds ratios indicating a substantial risk increase (OR = 1.60 for melanoma, 95% CI = 1.31-1.96) (ref: Gibson doi.org/10.1038/s41591-024-02837-7/). This highlights the importance of integrating genetic predisposition into treatment planning to mitigate future cancer risks. In another study, Subramanian et al. utilized a machine-learning framework to characterize the cellular ecosystems within soft tissue sarcomas, identifying 23 sarcoma-specific cell states that correlate with patient prognosis and response to immunotherapy (ref: Subramanian doi.org/10.1038/s43018-024-00743-y/). This approach underscores the potential of personalized therapies tailored to the unique genetic and cellular profiles of tumors. Furthermore, the exploration of threonine's role in glioblastoma by Wu et al. demonstrated how metabolic reprogramming can enhance protein translation in cancer stem cells, suggesting that targeting metabolic pathways may offer new therapeutic avenues (ref: Wu doi.org/10.1038/s43018-024-00748-7/).