The next-generation virtual cell: From spatiotemporal transcriptomic modeling to closed-loop target discovery in...

Drug discovery for complex diseases has long been constrained by high developmental costs and suboptimal clinical transition rates. To bridge the "translational gap" between in vitro target screening and in vivo therapeutic efficacy, life science and pharmacological research are undergoing a profound methodological paradigm shift: transitioning from traditional, static biochemical pathway models to "Next-Generation Virtual Cells" predicated on large-scale single-cell and spatial multi-omics data. However, the application of virtual cells should not rely solely on the unconstrained expansion of computational scale. Rather, it necessitates the precise deciphering of the biological context within complex diseases, including tissue-level spatial heterogeneity, multicellular communication networks, and the intricate tumor microenvironment (TME). This work reviews the recent advancements of next-generation virtual cells, focusing on their capacity for high-throughput resolution of drug mechanisms of action (MoA) through transcriptomic perturbation and the decoding of mechanisms underlying immune evasion and acquired resistance. We suggest that to dismantle the barriers between computational prediction and physiological response, it is necessary to deeply integrate in silico predictions with high-throughput wet-lab validation-such as organoid screening and cellular arrays-to establish a rigorous verification loop. By enhancing mechanistic interpretability and addressing the validation gap between transcriptomics and proteomics, the next generation of virtual cells is expected to accelerate the discovery of pharmacological targets and may provide new technological pathways for precision medicine.