An analysis of postmortem brain data found Alzheimer’s-related genetic signals that conventional comparisons of average gene activity missed. The single peer-reviewed study combined information from millions of cell nuclei and distinguished patterns across brain-cell populations. Its observations identify biological candidates for investigation; they do not establish a gene’s causal role or validate a test for living patients.
Genes with little difference in average activity still carried information associated with Alzheimer’s when analysed alongside other genes. A single peer-reviewed study of postmortem brain data found 27 gene–cell-type pairs supported by inherited-risk comparisons but missed by average-expression testing. The result distinguishes a gene’s contribution to a statistical model from how strongly its average activity changes.[1]
Researchers brought together approximately 4.2 million nuclei from 627 donors across six datasets. The analysis separated six major brain-cell populations and 41 subclusters. It combined multivariable prediction, gene-expression comparisons and inferred ordering of cellular states. Stress-related activity in astrocytes, cells supporting neurons, and signalling-related genes in neurons emerged among the convergent biological candidates.[1]
Patterns associated with clinical diagnosis and cognition did not always match those associated with brain pathology. Independent cohorts reproduced some signals. Because these are observations made after death, the inferred ordering cannot track disease progression directly in a living patient. Genetic overlap does not establish causation, and the analysis validates neither a clinical diagnostic test nor a gene-targeted treatment.[1]
