Mapping Alzheimer's neuropathology signatures to the whole brain transcriptome using machine learning data fusion

In plain English

The researchers learned disease-linked expression patterns where pathology data were available, then estimated where similar cellular patterns appear throughout the brain.

What they found

• The projected map highlighted region-, layer-, and cell-type-specific signatures associated with tau and beta-amyloid pathology.
• Reported vulnerable populations included neurons, astrocytes, and microglia across cortical and subcortical regions.

The catch

• The record is a preprint and has not completed peer review.
• This summary is based on the abstract; methods and supplementary analyses were not independently rechecked.
• The whole-brain results are model-based extrapolations rather than direct pathology measurements in every region.

Methods at a glance

Study type

Machine-learning data-fusion study

Biological system

Human Alzheimer's neuropathology and whole-brain transcriptomics

Organism

Human

Dataset

More than two million cortical cells from 427 people and three million atlas cells from 108 brain regions

Sample size

427 people in the cortical training data; 108 regions in the whole-brain atlas

Model architecture

Machine-learning transcriptomic estimators with cross-dataset data fusion

Training objective

Estimate neuropathology-associated transcriptomic signatures

Code

Not reported in the abstract

Data

The Siletti whole-human-brain atlas and cortical single-cell datasets are named at a collection level.

Tasks and metrics

Tasks: Pathology-signature estimation; Whole-brain extrapolation

Metrics: Ground-truth validation performance; Regional and cell-type signature localization

Baselines: Datasets with known regional ground truth

Principal findings

• The model produced region-, layer-, and cell-type-specific maps of pathology-associated transcriptomic signatures.

What remains unresolved

External validation: The abstract reports extensive validation on datasets with known ground truth

• The record is a preprint and has not completed peer review.
• This summary is based on the abstract; methods and supplementary analyses were not independently rechecked.
• Whole-brain outputs are extrapolated signatures rather than direct pathology measurements in every region.