Copy Number Variant Risk Scores Are Associated with Personalized Functional Brain Network Topography
Poster No:
683
Submission Type:
Abstract Submission
Authors:
Kevin Sun1, Zhiqiang Sha1, Benjamin Jung2, Joëlle Bagautdinova1, Laura Almasy2, Smrithi Prem1, Arielle Keller3, Michael Gandal1, Jakob Seidlitz2, Theodore Satterthwaite1, Aaron Alexander-Bloch1
Institutions:
1University of Pennsylvania, Philadelphia, PA, 2Children's Hospital of Philadelphia, Philadelphia, PA, 3University of Connecticut, Mansfield, CT
First Author:
Co-Author(s):
Introduction:
Large genetic deletions or duplications, known as copy number variants (CNVs), are major contributors to mental illness. While most CNVs are benign, individuals with CNVs often exhibit early-onset psychiatric symptoms (Stefansson et al., 2014). Based on gene annotations, CNV risk scores can quantify accumulative burden for either deletions or duplications. Our group and others have found associations among elevated CNV risk scores, lower cognitive ability, and higher psychopathology (Sha et al., 2024). Furthermore, CNVs are associated with total brain size and other deviations from normative brain structure (Alexander-Bloch et al., 2022; Modenato et al., 2021). Yet, how CNVs are related to functional brain organization remains understudied (Moreau et al., 2023). Here, we investigate the impact of CNV risk scores on personalized functional brain network (PFN) topography in the Adolescent Brain Cognitive Development (ABCD) Study℠.
Methods:
Genotyped data was acquired from the ABCD portal and CNVs were called using PennCNV and QuantiSNP, identifying 5760 CNVs across 8564 individuals. CNV risk scores were calculated based on gene annotations for loss-of-function intolerance (LOEUF). The cumulative inverse of LOEUF of all genes was computed for deletions (Del-iLOEUF) and duplications (Dup-iLOEUF) (Fig 1A).
Functional MRI data from the ABCD baseline (9-10-year-olds) were used to derive PFNs. Each participant's timeseries was decomposed through spatially constrained non-negative matrix factorization (Li et al., 2017), resulting in a PFN loading matrix of 17 networks across 59,412 cortical vertices (Fig 1B). Ridge regression models trained on PFN loading matrices were used to associate PFN topography to each CNV risk score (Fig 1C), using two-fold cross-validation across matched split-half subsets. Sensitivity analyses were performed with estimated total intracranial volume (eTIV) as a covariate. Haufe-transformed regression weights (Haufe et al., 2014) were analyzed to identify cortical regions and networks driving these associations.
Functional MRI data from the ABCD baseline (9-10-year-olds) were used to derive PFNs. Each participant's timeseries was decomposed through spatially constrained non-negative matrix factorization (Li et al., 2017), resulting in a PFN loading matrix of 17 networks across 59,412 cortical vertices (Fig 1B). Ridge regression models trained on PFN loading matrices were used to associate PFN topography to each CNV risk score (Fig 1C), using two-fold cross-validation across matched split-half subsets. Sensitivity analyses were performed with estimated total intracranial volume (eTIV) as a covariate. Haufe-transformed regression weights (Haufe et al., 2014) were analyzed to identify cortical regions and networks driving these associations.
Results:
PFN topography was significantly associated with Del-iLOEUF (Split-Half-A: r=0.20, p<0.001; Split-Half-B: r=0.20, p<0.001) (Fig 2A) and Dup-iLOEUF (Split-Half-A: r=0.06, p<0.01; Split-Half-B: r=0.07, p<0.01) (Fig 2B). Including eTIV as a covariate returned significant results: Del-iLOEUF (Split-Half-A: r=0.17, p<0.001; Split-Half-B: r=0.17, p<0.001), Dup-iLOEUF (Split-Half-A: r=0.05, p<0.01; Split-Half-B: r=0.06, p<0.01).
To identify important cortical regions driving the association between PFN topography and CNV risk scores, we mapped the magnitude of summed weights across networks onto the cortical surface for Del-iLOEUF (Fig 2C) and Dup-iLOEUF (Fig 2D). We found a significant spatial correlation between weight maps of Del-iLOEUF and Dup-iLOEUF (r = 0.67, pspin= 0.001), implying that shared biological pathways drive functional brain organization in the context of deletions and duplications.
The top 1% contributing regions (Fig 2E, F) overlapped between Del-iLOEUF and Dup-iLOEUF, namely bilateral superior parietal lobule and left superior temporal gyrus. Directional network topography analysis (Fig 2G, H) highlighted CNV influences on association networks (Fig 2I), particularly dorsal attention networks (networks 14 and 5). Indeed, superior parietal lobule network 14 topography was associated with higher risk scores, whereas network 5 topography at the same region was associated with lower risk scores (Fig 2J).
To identify important cortical regions driving the association between PFN topography and CNV risk scores, we mapped the magnitude of summed weights across networks onto the cortical surface for Del-iLOEUF (Fig 2C) and Dup-iLOEUF (Fig 2D). We found a significant spatial correlation between weight maps of Del-iLOEUF and Dup-iLOEUF (r = 0.67, pspin= 0.001), implying that shared biological pathways drive functional brain organization in the context of deletions and duplications.
The top 1% contributing regions (Fig 2E, F) overlapped between Del-iLOEUF and Dup-iLOEUF, namely bilateral superior parietal lobule and left superior temporal gyrus. Directional network topography analysis (Fig 2G, H) highlighted CNV influences on association networks (Fig 2I), particularly dorsal attention networks (networks 14 and 5). Indeed, superior parietal lobule network 14 topography was associated with higher risk scores, whereas network 5 topography at the same region was associated with lower risk scores (Fig 2J).
Conclusions:
PFN topography is associated with CNV risk scores for both deletions and duplications beyond the effect of total brain size. Shared cortical regions and network topographies suggest that biological pathways downstream of deletions and duplications overlap. In particular, dorsal attention network topography was altered in youth with elevated CNV risk scores, a possible intermediate phenotype for behavioral attention problems observed in youth with CNVs (Sha et al., 2024).
Disorders of the Nervous System:
Neurodevelopmental/ Early Life (eg. ADHD, autism)
Genetics:
Genetic Association Studies 1
Lifespan Development:
Early life, Adolescence, Aging
Modeling and Analysis Methods:
fMRI Connectivity and Network Modeling 2
Multivariate Approaches
Keywords:
Computational Neuroscience
Cortex
Development
FUNCTIONAL MRI
Machine Learning
Multivariate
Other - Genetics
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Are you Internal Review Board (IRB) certified? Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
2. Brownstein, C. A., et al. (2022). Similar Rates of Deleterious Copy Number Variants in Early-Onset Psychosis and Autism Spectrum Disorder. American Journal of Psychiatry, 179(11), 853–861. https://doi.org/10.1176/appi.ajp.21111175
3. Haufe, S., et al. (2014). On the interpretation of weight vectors of linear models in multivariate neuroimaging. NeuroImage, 87, 96–110. https://doi.org/10.1016/j.neuroimage.2013.10.067
4. Keller, A. S., et al. (2024). A general exposome factor explains individual differences in functional brain network topography and cognition in youth. Developmental Cognitive Neuroscience, 66. https://doi.org/10.1016/j.dcn.2024.101370
5. Li, H., et al. (2017). Large-scale sparse functional networks from resting state fMRI. NeuroImage, 156, 1–13. https://doi.org/10.1016/j.neuroimage.2017.05.004
6. Modenato, C., et al. (2021). Effects of eight neuropsychiatric copy number variants on human brain structure. Translational Psychiatry, 11(1). https://doi.org/10.1038/s41398-021-01490-9
7. Moreau, C. A., et al. (2023). Genetic Heterogeneity Shapes Brain Connectivity in Psychiatry. Biological Psychiatry, 93(1), 45–58. https://doi.org/10.1016/j.biopsych.2022.08.024
8. Sha, Z., et al. (2024). The copy number variant architecture of psychopathology and cognitive development in the ABCD® study. medRxiv https://doi.org/10.1101/2024.05.14.24307376
9. Stefansson, H., et al. (2014). CNVs conferring risk of autism or schizophrenia affect cognition in controls. Nature, 505(7483), 361–366. https://doi.org/10.1038/nature12818