AI research identifies key mechanisms in glioblastoma and preterm brain development

Two separate studies utilizing artificial intelligence have provided new insights into complex biological processes in glioblastoma and neonatal brain development.

In glioblastoma research, a new AI framework called scDORI has been used to analyze multi-omics data from over one million patient-derived cells. The research reveals that glioblastoma cell states follow fixed, hierarchical transition trajectories rather than changing chaotically. Researchers identified the regulator MYT1L, which, when overexpressed, reduces cellular plasticity and slows tumor growth in animal models. This finding suggests a shift in treatment focus toward cellular reprogramming to bypass therapy resistance.

In neonatal research, a study published in Science indicates that preterm birth alters the biological development of the cerebellum rather than simply delaying it. By combining neuroimaging of 176 preterm infants with spatial transcriptomics, researchers found that the cerebellum loses volume weekly when born outside the protected womb. The findings suggest that preterm infants undergo a different molecular programming of cell types, supporting a shift in neonatal intensive care toward prioritizing developmental maturity over chronological age.