SPACER AI maps signals linked to T-cell recruitment

Researchers at UT MD Anderson and collaborating institutions developed SPACER, an interpretable AI model that analyzes spatial gene-activity data to identify signals associated with immune-cell recruitment in tissue. The team applied it to 37 datasets across several cancer types, according to the center’s summary of the study. ^[2]

The analysis linked tumor gene activity involved in antigen presentation with greater T-cell recruitment, while higher mucin levels were associated with reduced T-cell engagement; the researchers suggest mucins may create a physical barrier. These are associations from data analysis, not proof that mucins cause immune exclusion or that the model can guide treatment. The institutional summary says the findings appeared in Nature Biomedical Engineering, but the journal article could not be independently accessed for this verification. ^[2]