Overview
This document serves as a formal correction to the 2019 publication by Mathys et al., "Single-cell transcriptomic analysis of Alzheimer’s disease" (Nature 570, 332–337). The correction was issued to address specific administrative inaccuracies regarding funding and, more critically, to rectify a data labeling error within the supplementary materials that impacts the interpretation of differential expression across disease stages.
Data Correction: Supplementary Table 2
The primary technical correction involves Supplementary Table 2, which details differential expression across various cell types. In the original version of the paper, the mean expression values for different pathological stages were inadvertently swapped in the supplementary data files.
Affected Cell Types
The labeling error occurred across multiple worksheets in the supplementary file, representing a broad range of brain cell populations:
- Excitatory neurons (Ex)
- Inhibitory neurons (In)
- Astrocytes (Ast)
- Oligodendrocytes (Oli)
- Oligodendrocyte precursor cells (Opc)
- Microglia (Mic)
The Specific Error
Under the heading ‘early-pathology vs late-pathology differential expression,’ the labels for columns Y and Z were reversed. Specifically, the data originally labeled as early.pathology.mean actually represented the late.pathology.mean, and vice versa. This correction ensures that researchers utilizing these datasets for meta-analyses or validation studies are comparing the correct mean expression values for early-stage versus late-stage Alzheimer’s pathology.
Administrative and Funding Updates
In addition to the data corrections, the authors have updated the acknowledgments and funding sections to ensure full transparency and accuracy:
- Funding Recognition: The Cure Alzheimer’s Fund (CAF) has been added to the acknowledgments.
- Grant Correction: The NIH grant number originally listed as ‘RF1AG054012’ has been corrected to ‘U01MH119509’.
Significance
Maintaining the integrity of large-scale transcriptomic datasets is vital for the scientific community. While the core conclusions of the original paper remain supported, this correction is essential for researchers performing secondary analyses. By fixing the column swap in Supplementary Table 2, the authors ensure that the comparative expression profiles accurately reflect the progression of Alzheimer’s disease pathology. The online version of the original article has been updated to reflect these changes.
Summary updated