Back

A cross-disorder analysis of CNVs finds novel loci and dose-dependent relationships of genes to psychiatric traits

Shanta, O.; Klein, M.; Sacks, M.; MacDonald, J. R.; Maihofer, A. X.; Ahangari, M.; Engchuan, W.; Thiruvahindrapuram, B.; Guevara, J.; Hong, O.; Huguet, G.; Sonderby, I.; Kalyuzhny, M.; Adams, M. J.; Adolfsson, R.; Agartz, I.; Aiello, A. E.; Alda, M.; Allardyce, J.; Amstadter, A. B.; Andlauer, T. F. M.; Andreassen, O. A.; Artigas, M. S.; Austin, S. B.; Ayub, M.; Baker, D. G.; Bass, N.; Baune, B. T.; Bayas, M.; Berger, K.; Biernacka, J. M.; Bigdeli, T.; Bisson, J. I.; Blackwood, D.; Boks, M.; Braff, D.; Bramon, E.; Breen, G.; Brueckl, T.; Bryant, R. A.; Bulik, C. M.; Buxbaum, J.; Cairns, M. J.;

2025-07-15 genetic and genomic medicine
10.1101/2025.07.11.25331310 medRxiv
Show abstract

Rare copy number variants (CNVs) are a key component of the genetic basis of psychiatric conditions, but have not been well characterized for most. We conducted a genome-wide CNV analysis across six diagnostic categories (N = 574,965): autism (ASD), ADHD, bipolar disorder (BD), major depressive disorder (MDD), PTSD, and schizophrenia (SCZ). We identified 35 genome-wide significant associations at 18 loci, including novel associations in SCZ (SMYD3, USP7-HAPSTR1) and in the combined cross-disorder analysis (ASTN2). Rare CNVs accounted for 1-3% of heritability across diagnoses. In ASD, associations were uniformly positive, consistent with autism having diverse etiologies and clinical presentations. By contrast, CNVs showed a dose-dependent relationship for other diagnoses, including SCZ and PTSD, with reciprocal deletions and duplications having inversely correlated effects and distinct genotype-phenotype relationships. Our findings suggest that genes have effects that are both dose-dependent and pleiotropic, such that a positive influence on one dimension of psychopathology may be accompanied by positive or negative effects on others.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.