An Explainable MRI-Conditioned Generative Framework for Gene-Feature Representations in Alzheimer’s Disease
- 1Department of Biostatistics & Health Data Science, Indiana University School of Medicine, Indianapolis, IN 46202, USA
- 2Department of Computer Science, Indiana University Bloomington, Bloomington, IN 47408, USA
- 3Department of Mathematics, Indiana University School of Science, Indianapolis, IN 46202, USA
Abstract
Magnetic resonance imaging (MRI) and peripheral-blood transcriptomics capture complementary aspects of Alzheimer’s disease (AD), but whether MRI can condition a meaningful molecular representation remains unclear. We developed and independently evaluated an explainable imaging-genetics conditional variational autoencoder (IG-CVAE) in Alzheimer’s Disease Neuroimaging Initiative (ADNI) and AddNeuroMed (ANM) cohorts using fixed 80% train-validation and 20% independent test sets with five-fold cross-validation. IG-CVAE comprises a genetic branch encoding expression into 128-dimensional gene-feature vectors (GFV128), an imaging branch encoding MRI into 128-dimensional MRI-feature vectors (MFV128), and a conditional variational autoencoder that generates GFV128 from MFV128. Paired analyses included 289 ADNI and 98 ANM participants. Test area under the receiver operating characteristic curve (ROC-AUC) values in ADNI/ANM were 0.646/0.851 for the genetic branch, 0.942/0.752 for the imaging branch, and 0.939/0.869 for generated GFV128. Generated-observed GFV128 agreement was moderate (cosine similarity, 0.590/0.589; coefficient of determination (R2), 0.161/0.211). Molecular importance was cohort-specific at the gene level, with no significant pathway-level convergence, whereas MRI perturbation showed cross-cohort convergence in cerebral white matter linked to distinct disease-informative molecular features. Overall, structural MRI retained disease-relevant information sufficient to generate disease-informative molecular representations, supporting IG-CVAE as a participant-paired hypothesis-generating imaging–transcriptomic framework while providing only limited evidence of participant-specific molecular correspondence without implying raw transcriptome reconstruction. The ANM test estimates should be interpreted cautiously given the small paired independent test set (n = 20).
Related
- A diagnostic BRICS fragment-connection framework for infrared-conditioned molecular reconstruction
- Subtype-Conditioned NRF2-KEAP1 Signaling in Breast Cancer: A Narrative Review and Hypothesis-Generating Framework from Chemoprevention to Therapy Resistance.
- KAT-Net: a hyperbolic routing framework for joint lysine acylation site prediction with shared sequence–structure representations
- GenAI-Net: A generative AI framework for automated biomolecular network design
- Explainable computational framework for subtype classification and biomarker identification in head and neck squamous cell carcinoma using machine learning
BibTeX
@article{huang2026explainable,
title = {{An Explainable MRI-Conditioned Generative Framework for Gene-Feature Representations in Alzheimer’s Disease}},
author = {Xiaoqing Huang and Md Hasan Imam Shihab and Ronakkumar Laxmanbhai Bhagchandani and Yuxin Dong and Yijie Wang},
journal = {International Journal of Molecular Sciences},
year = {2026},
month = oct,
doi = {10.3390/ijms27219741},
url = {https://www.mdpi.com/1422-0067/27/21/9741},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - An Explainable MRI-Conditioned Generative Framework for Gene-Feature Representations in Alzheimer’s Disease
AU - Xiaoqing Huang
AU - Md Hasan Imam Shihab
AU - Ronakkumar Laxmanbhai Bhagchandani
AU - Yuxin Dong
AU - Yijie Wang
PY - 2026
DA - 2026/10/30
JO - International Journal of Molecular Sciences
DO - 10.3390/ijms27219741
UR - https://www.mdpi.com/1422-0067/27/21/9741
AB - Magnetic resonance imaging (MRI) and peripheral-blood transcriptomics capture complementary aspects of Alzheimer’s disease (AD), but whether MRI can condition a meaningful molecular representation remains unclear. We developed and independently evaluated an explainable imaging-genetics conditional variational autoencoder (IG-CVAE) in Alzheimer’s Disease Neuroimaging Initiative (ADNI) and AddNeuroMed (ANM) cohorts using fixed 80% train-validation and 20% independent test sets with five-fold cross-validation. IG-CVAE comprises a genetic branch encoding expression into 128-dimensional gene-feature vectors (GFV128), an imaging branch encoding MRI into 128-dimensional MRI-feature vectors (MFV128), and a conditional variational autoencoder that generates GFV128 from MFV128. Paired analyses included 289 ADNI and 98 ANM participants. Test area under the receiver operating characteristic curve (ROC-AUC) values in ADNI/ANM were 0.646/0.851 for the genetic branch, 0.942/0.752 for the imaging branch, and 0.939/0.869 for generated GFV128. Generated-observed GFV128 agreement was moderate (cosine similarity, 0.590/0.589; coefficient of determination (R2), 0.161/0.211). Molecular importance was cohort-specific at the gene level, with no significant pathway-level convergence, whereas MRI perturbation showed cross-cohort convergence in cerebral white matter linked to distinct disease-informative molecular features. Overall, structural MRI retained disease-relevant information sufficient to generate disease-informative molecular representations, supporting IG-CVAE as a participant-paired hypothesis-generating imaging–transcriptomic framework while providing only limited evidence of participant-specific molecular correspondence without implying raw transcriptome reconstruction. The ANM test estimates should be interpreted cautiously given the small paired independent test set (n = 20).
SN - 1422-0067
ER -