Residual genetic connectivity beyond shared liability in 14 psychiatric disorders

Area of Science:

• Psychiatric Genomics
• Behavioral Genetics
• Computational Psychiatry

Background:

• Transdiagnostic genetic factor models aim to explain shared liability across psychiatric disorders.
• These models may not fully account for systematic pairwise genetic overlap between specific disorders.

Purpose of the Study:

• To investigate unexplained pairwise genetic overlap in psychiatric disorders after accounting for shared liability.
• To identify specific disorder pairs with significant residual genetic correlations and analyze their structural patterns.

Main Methods:

• Utilized publicly available Psychiatric Genomics Consortium cross-disorder genetic correlations.
• Computed edge-level residual genetic correlations by subtracting model-implied correlations from observed correlations.
• Employed Monte Carlo sampling to propagate uncertainty and identified residual hubs.

Main Results:

• Residual genetic structure was concentrated in specific disorders, not uniformly distributed.
• Obsessive-compulsive disorder (OCD)-anxiety showed the largest positive residual edge.
• OCD and attention-deficit/hyperactivity disorder (ADHD) emerged as significant residual hubs.

Conclusions:

• Higher-order genetic factors capture broad liability but leave structured residual associations between specific disorder pairs.
• Examining residual patterns clarifies the limitations of current transdiagnostic architectures.
• Residual analysis can inform more explainable genetic models for psychiatric disorders.