Genomic and clinical correlates of 5 deep learning‐based neuroimaging signatures of neurodegeneration, evaluated in ADSP
Bibliographic record
Abstract
BACKGROUND: Deep learning methods help to disentangle heterogeneity of brain aging and find distinct neuroimaging patterns of neurodegeneration. However, the relationship between aging-related brain atrophy patterns and genetics is complex and requires further exploration. METHOD: Patterns of regional bran volumes extracted from T1 scans and Whole-genome sequencing (WGS) data in The Alzheimer's Disease Sequencing Project (ADSP) were analyzed (2401 subjects, age = 72.75 ± 9.08; 54.07% female). We investigated 5 recently published brain aging patterns (R-indices) (Yang et al., 2024), and evaluated out-of-sample reproducibility of previously reported associations between R-indices and diagnostic groups (CN vs MCI/AD). We conducted genome-wide association analysis to investigate genetic associations of R-indices controlling for confounders (e.g., age). RESULT: Linear regression models identified significant group differences with large effect size in diagnostic groups in R2 (Cohen's d=1.05, p = 1.66e-105; this is a typical AD pattern of atrophy) and R3 (d=0.81, p = 2.16e-66), and moderate effect size in R5 (d=0.44, p = 9.59e-22; this pattern was previously associated with a variety of cardiovascular risk factors and immune system markers). Genetic analyses identified 5 genes significantly associated with R1, R2, R3 and R5 indices (Figure 1). R1, characterized with subcortical atrophy, had associations with A1CF (chr10), which has been associated with urate levels, gout and colorectal cancer. R2, with focal medial temporal lobe atrophy, was associated with EXT1 (chr8), which has been associated with BMI, general cognitive ability, cortical thickness and insomnia. R3, with parieto-temporal atrophy patterns, was associated with CUL4A (chr13), which has been linked to blood cell volume/distribution, bipolar disorder, HDL cholesterol, corpuscular hemoglobin and atrial fibrillation. R5, with primarily perisylvian atrophy, was associated with DDX39B and ATP6V1G2-DDX39B (chr6), which have been associated with insomnia, general cognitive ability and blood cell measures. CONCLUSION: We evaluated 5 recently established brain aging indices, which capture the heterogeneity of structural brain aging, in ADSP participants. R-indices replicated previously reported associations with AD groups (Yang et al., 2024), indicating the generalizability of the model. We identified different associations with genes that were previously linked to various traits. These findings provide new insights into the exploration of heterogeneity of neurodegeneration and related genetic risk factors.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".