Transcriptomics-Driven Biomarkers for Major Depressive Disorder (MDD)

Project Overview

Major Depressive Disorder (MDD) is a severe psychiatric condition with complex functional and molecular mechanisms. Traditional structural imaging biomarkers have limited reproducibility and sensitivity — functional biomarkers like regional homogeneity (ReHo) demonstrate much stronger effects in large cohort studies.

This project adopts a transcriptomics-driven multi-modal framework to link RNA-seq–derived gene expression signatures to functional neuroimaging patterns in MDD, providing mechanistic insights and high-value biological interpretation for computational psychiatry.

Publication Reference:
Multimodal Neuroimaging and Transcriptomic Correlates of Major Depressive Disorder – JAMA Psychiatry

🔑 Key Scientific Insights

• Functional brain alterations outperform structural markers:
  ReHo-based functional MRI deficits show 2–3× larger effect sizes than cortical thickness reductions, confirming that MDD is primarily a disorder of functional dysregulation rather than structural atrophy.

• Robust hypoperfusion patterns:
  ReHo deficits strongly co-localize with regionally specific reductions in cerebral blood flow (RCBF), indicating highly reproducible cortical hypoperfusion patterns.

• Regional Vulnerability Index (RVI):
  Functional RVI (ReHo-derived) shows stronger and more consistent effects than any individual brain region and outperforms structural RVI measures.

• Large-scale cross-dataset validation:
  Findings were replicated across 15,501 participants from UK Biobank, ENIGMA, Amish Connectome Project, and independent clinical cohorts, confirming robustness and generalizability.

• Biological interpretation via transcriptomics:
  Integration of ReHo/RCBF with RNA-seq–based gene expression data reveals molecular mechanisms underlying functional deficits, highlighting synaptic, inflammatory, and neurovascular pathways.

🧬 Transcriptomics & Molecular Integration

• Brain-region–specific RNA-seq profiles mapped to functional deficits
• Differential expression of key synaptic, inflammatory, and neurovascular genes
• Pathway enrichment analyses (GO/KEGG) tied to neurofunctional alterations
• Integration of transcriptomic gradients with cortical maps for mechanistic insight

This positioning emphasizes transcriptomics as a central pillar, making the project highly relevant for researchers in multi-omics, systems biology, and computational psychiatry.

🔬 Functional Neuroimaging Biomarkers

• Regional Homogeneity (ReHo): voxel-level local synchrony in resting-state fMRI
• Regional Cerebral Blood Flow (RCBF): perfusion measures aligned with ReHo deficits
• Functional deficits in MDD consistently exceed structural cortical thickness changes

🔁 Workflow Schematic

The schematic illustrates the multi-modal workflow:

1. RNA-seq preprocessing and transcriptomic feature extraction
2. Functional MRI preprocessing and ReHo/RCBF calculation
3. Linking transcriptomic profiles with functional imaging deficits
4. Integrating multi-modal features for biomarker discovery
5. Mechanistic interpretation and predictive modeling

This framework provides biologically interpretable, cross-modal biomarkers suitable for diagnostics, stratification, and personalized interventions.

📌 Highlights (Summary)

• Functional deficits in MDD > Structural cortical reductions
• ReHo-based RVI shows strong reproducibility across datasets
• Transcriptomic integration links imaging phenotypes to molecular mechanisms
• Cross-validated over 15,501 participants
• Supports precision psychiatry approaches with multi-omics biomarkers

👥 Authorship & Credits

This project is based on a peer-reviewed research article published in JAMA Psychiatry. The present page emphasizes transcriptomics-informed modeling and computational extensions of the original findings.

Project Contributors:

• Peter Kochunov, PhD
  Professor of Psychiatry & Neuroimaging
  Senior author, leading neuroimaging-genomics integration and transcriptomic interpretation.

• Syed Ibrar Hussain, PhD
  Department of Mathematics, University of Houston, Houston, TX, USA Co-author of the original publication
  Contributed to transcriptomics analysis, functional neuroimaging integration, and computational modeling.

• Additional Co-authors
  Contributed to study design, data acquisition, transcriptomic profiling, neuroimaging analysis, and statistical modeling.

🤝 Collaboration & Contact

Open for collaboration in:

• Transcriptomics and RNA-seq integration
• Computational psychiatry and machine learning
• Multi-omics biomarker discovery
• Functional neuroimaging research

Contact via email or LinkedIn to discuss potential collaborations.