Mapping Mutual Information of Neuroimages and Allelic Variants Unveils Determinants of Brain Aging
Poster No:
866
Submission Type:
Abstract Submission
Authors:
Nicholas Kim1, Anar Amgalan1, Nahian Chowdhury1, Nikhil Chaudhari1, Phoebe Imms1, Paul Thompson2, Andrei Irimia1
Institutions:
1University of Southern California, Los Angeles, CA, 2Imaging Genetics Center, Keck School of Medicine of University of Southern California, Los Angeles, CA
First Author:
Co-Author(s):
Paul Thompson, PhD
Imaging Genetics Center, Keck School of Medicine of University of Southern California
Los Angeles, CA
Imaging Genetics Center, Keck School of Medicine of University of Southern California
Los Angeles, CA
Introduction:
Single-nucleotide polymorphisms (SNPs) are genetic variations with effects on brain aging. Mapping their influence on brain structure can elucidate the genetic correlates of neurological/psychiatric disorders. However, most genome-wide association studies use linear measures of dependents which ignore the nonlinear dependence of phenotype on genotype. In this study, we used mutual information (MI), an information-theoretic measure of reciprocal (nonlinear) dependence between two variables, to map how SNPs act nonlinearly on cortical morphology.
Methods:
In 6,000 UK Biobank participants (including 3,209 females) aged 54 to 84, we extracted mean T1-weighted MRI intensities at each cortical location. Intensities were projected onto a reference atlas to accommodate variations in brain morphology. We mapped the MI between MRI intensity and 706,000 SNP variants to quantify the dose-response relationship between (A) SNP variant number and (B) the location-dependent MRI intensity of the cerebral cortex.
A cortical map of MI with MRI intensity was generated for each SNP. The mean MI of each map was used to rank SNPs according to genetic influence. Maps were normalized by calculating z-scores at each cortical location using the distribution of SNP MIs at that specific location.
A cortical map of MI with MRI intensity was generated for each SNP. The mean MI of each map was used to rank SNPs according to genetic influence. Maps were normalized by calculating z-scores at each cortical location using the distribution of SNP MIs at that specific location.
Results:
Cortical MI maps reveal novel genetic influences on brain structure that involve neurocognitive and psychiatric conditions. SNPs found to act most strongly on cortical structure involved the introns of PRPF31 (previously linked to smaller brain volume), NHSl1 (cortical surface area), and LINC00299 (microglia; activity in early neurodegeneration).
KIF6, linked to attention-deficit/hyperactivity disorder (ADHD), ranks among the SNPs with strongest effect on MRI intensity. Prefrontal cortex mediates behavioral inhibition frequently affected in ADHD. As depicted in Figure 1, the number of KIF6 alleles has significantly higher MI (p < 0.05) with MRI intensity than the average SNP in this lobe, specifically in the right ventrolateral region. Such influence is notably pronounced in features of the prefrontal cortex in the right hemisphere, a notable lateralization effect consistent with findings in prior studies (Arnsten, 2009).
The APOE allele, the strongest genetic risk factor for Alzheimer's disease (AD), ranks among the top 0.1% of SNPs acting on MRI intensity. As shown in Figure 2, the number of APOE-ε4 alleles has significantly higher MI (p < 0.05) with MRI intensity in the medial parietal lobe, temporal lobes, and left inferior occipital lobe, three brain regions responsible for long-term memory recall, language processing, and visual perception (Ackerman, 1992). These findings reflect AD symptoms, which include memory loss, impaired verbal fluency, and reduced peripheral vision (Weintraub et al., 2012).
Other findings reveal similar relationships between SNPs involved in neurological/psychiatric conditions and brain structure. Results were reproduced in a larger sample of 43,000 UKBB participants.
KIF6, linked to attention-deficit/hyperactivity disorder (ADHD), ranks among the SNPs with strongest effect on MRI intensity. Prefrontal cortex mediates behavioral inhibition frequently affected in ADHD. As depicted in Figure 1, the number of KIF6 alleles has significantly higher MI (p < 0.05) with MRI intensity than the average SNP in this lobe, specifically in the right ventrolateral region. Such influence is notably pronounced in features of the prefrontal cortex in the right hemisphere, a notable lateralization effect consistent with findings in prior studies (Arnsten, 2009).
The APOE allele, the strongest genetic risk factor for Alzheimer's disease (AD), ranks among the top 0.1% of SNPs acting on MRI intensity. As shown in Figure 2, the number of APOE-ε4 alleles has significantly higher MI (p < 0.05) with MRI intensity in the medial parietal lobe, temporal lobes, and left inferior occipital lobe, three brain regions responsible for long-term memory recall, language processing, and visual perception (Ackerman, 1992). These findings reflect AD symptoms, which include memory loss, impaired verbal fluency, and reduced peripheral vision (Weintraub et al., 2012).
Other findings reveal similar relationships between SNPs involved in neurological/psychiatric conditions and brain structure. Results were reproduced in a larger sample of 43,000 UKBB participants.
Conclusions:
Our findings suggest that mapping MI between MRI intensities and SNPs can reveal the genetic influences of SNPs on brain structure in health and in neurological/psychiatric disease.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 2
Psychiatric (eg. Depression, Anxiety, Schizophrenia)
Genetics:
Genetic Modeling and Analysis Methods 1
Modeling and Analysis Methods:
Univariate Modeling
Keywords:
Aging
Attention Deficit Disorder
Computational Neuroscience
Cortex
Data analysis
Degenerative Disease
MRI
Neurological
Statistical Methods
STRUCTURAL MRI
1|2Indicates the priority used for review
Provide references using author date format
Arnsten, A. F. T. (2009), 'The Emerging Neurobiology of Attention Deficit Hyperactivity Disorder: The Key Role of the Prefrontal Association Cortex', The Journal of Pediatrics, vol. 154, no. 5, I-S43
Weintraub, S. (2012), 'The Neuropsychological Profile of Alzheimer Disease', Cold Spring Harbor Perspectives in Medicine, vol. 2, no. 4, a006171