Longitudinal genotype-phenotype association study through temporal structure auto-learning predictive model

Xiaoqian Wang, Jingwen Yan, Xiaohui Yao, Sungeun Kim, Kwangsik Nho, Shannon L. Risacher, Andrew Saykin, Li Shen, Heng Huang

Research output: Contribution to journalArticle

Abstract

With the rapid development of high-throughput genotyping and neuroimaging techniques, imaging genetics has drawn significant attention in the study of complex brain diseases such as Alzheimer's disease (AD). Research on the associations between genotype and phenotype improves the understanding of the genetic basis and biological mechanisms of brain structure and function. AD is a progressive neurodegenerative disease; therefore, the study on the relationship between single nucleotide polymorphism (SNP) and longitudinal variations of neuroimaging phenotype is crucial. Although some machine learning models have recently been proposed to capture longitudinal patterns in genotype-phenotype association studies, most machine-learning models base the learning on fixed structure among longitudinal prediction tasks rather than automatically learning the interrelationships. In response to this challenge, we propose a new automated time structure learning model to automatically reveal the longitudinal genotype-phenotype interactions and exploits such learned structure to enhance the phenotypic predictions. We proposed an efficient optimization algorithm for our model and provided rigorous theoretical convergence proof. We performed experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort for longitudinal phenotype prediction, including 3123 SNPs and 2 biomarkers (Voxel-Based Morphometry and FreeSurfer). The empirical results validate that our proposed model is superior to the counterparts. In addition, the best SNPs identified by our model have been replicated in the literature, which justifies our prediction.

Original languageEnglish (US)
Pages (from-to)809-824
Number of pages16
JournalJournal of Computational Biology
Volume25
Issue number7
DOIs
StatePublished - Jul 1 2018

Fingerprint

Predictive Model
Genetic Association Studies
Genotype
Neuroimaging
Phenotype
Single Nucleotide Polymorphism
Alzheimer Disease
Association reactions
Learning
Alzheimer's Disease
Genotyping Techniques
Prediction
Brain Diseases
Neurodegenerative Diseases
Learning systems
Brain
Machine Learning
Biomarkers
Model
Morphometry

Keywords

  • Alzheimer's disease
  • genotype-phenotype association prediction
  • longitudinal study
  • low-rank model
  • temporal structure auto-learning

ASJC Scopus subject areas

  • Modeling and Simulation
  • Molecular Biology
  • Genetics
  • Computational Mathematics
  • Computational Theory and Mathematics

Cite this

Longitudinal genotype-phenotype association study through temporal structure auto-learning predictive model. / Wang, Xiaoqian; Yan, Jingwen; Yao, Xiaohui; Kim, Sungeun; Nho, Kwangsik; Risacher, Shannon L.; Saykin, Andrew; Shen, Li; Huang, Heng.

In: Journal of Computational Biology, Vol. 25, No. 7, 01.07.2018, p. 809-824.

Research output: Contribution to journalArticle

Wang, Xiaoqian ; Yan, Jingwen ; Yao, Xiaohui ; Kim, Sungeun ; Nho, Kwangsik ; Risacher, Shannon L. ; Saykin, Andrew ; Shen, Li ; Huang, Heng. / Longitudinal genotype-phenotype association study through temporal structure auto-learning predictive model. In: Journal of Computational Biology. 2018 ; Vol. 25, No. 7. pp. 809-824.
@article{87835a0e057e4ee69aa4b28b4f904374,
title = "Longitudinal genotype-phenotype association study through temporal structure auto-learning predictive model",
abstract = "With the rapid development of high-throughput genotyping and neuroimaging techniques, imaging genetics has drawn significant attention in the study of complex brain diseases such as Alzheimer's disease (AD). Research on the associations between genotype and phenotype improves the understanding of the genetic basis and biological mechanisms of brain structure and function. AD is a progressive neurodegenerative disease; therefore, the study on the relationship between single nucleotide polymorphism (SNP) and longitudinal variations of neuroimaging phenotype is crucial. Although some machine learning models have recently been proposed to capture longitudinal patterns in genotype-phenotype association studies, most machine-learning models base the learning on fixed structure among longitudinal prediction tasks rather than automatically learning the interrelationships. In response to this challenge, we propose a new automated time structure learning model to automatically reveal the longitudinal genotype-phenotype interactions and exploits such learned structure to enhance the phenotypic predictions. We proposed an efficient optimization algorithm for our model and provided rigorous theoretical convergence proof. We performed experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort for longitudinal phenotype prediction, including 3123 SNPs and 2 biomarkers (Voxel-Based Morphometry and FreeSurfer). The empirical results validate that our proposed model is superior to the counterparts. In addition, the best SNPs identified by our model have been replicated in the literature, which justifies our prediction.",
keywords = "Alzheimer's disease, genotype-phenotype association prediction, longitudinal study, low-rank model, temporal structure auto-learning",
author = "Xiaoqian Wang and Jingwen Yan and Xiaohui Yao and Sungeun Kim and Kwangsik Nho and Risacher, {Shannon L.} and Andrew Saykin and Li Shen and Heng Huang",
year = "2018",
month = "7",
day = "1",
doi = "10.1089/cmb.2018.0008",
language = "English (US)",
volume = "25",
pages = "809--824",
journal = "Journal of Computational Biology",
issn = "1066-5277",
publisher = "Mary Ann Liebert Inc.",
number = "7",

}

TY - JOUR

T1 - Longitudinal genotype-phenotype association study through temporal structure auto-learning predictive model

AU - Wang, Xiaoqian

AU - Yan, Jingwen

AU - Yao, Xiaohui

AU - Kim, Sungeun

AU - Nho, Kwangsik

AU - Risacher, Shannon L.

AU - Saykin, Andrew

AU - Shen, Li

AU - Huang, Heng

PY - 2018/7/1

Y1 - 2018/7/1

N2 - With the rapid development of high-throughput genotyping and neuroimaging techniques, imaging genetics has drawn significant attention in the study of complex brain diseases such as Alzheimer's disease (AD). Research on the associations between genotype and phenotype improves the understanding of the genetic basis and biological mechanisms of brain structure and function. AD is a progressive neurodegenerative disease; therefore, the study on the relationship between single nucleotide polymorphism (SNP) and longitudinal variations of neuroimaging phenotype is crucial. Although some machine learning models have recently been proposed to capture longitudinal patterns in genotype-phenotype association studies, most machine-learning models base the learning on fixed structure among longitudinal prediction tasks rather than automatically learning the interrelationships. In response to this challenge, we propose a new automated time structure learning model to automatically reveal the longitudinal genotype-phenotype interactions and exploits such learned structure to enhance the phenotypic predictions. We proposed an efficient optimization algorithm for our model and provided rigorous theoretical convergence proof. We performed experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort for longitudinal phenotype prediction, including 3123 SNPs and 2 biomarkers (Voxel-Based Morphometry and FreeSurfer). The empirical results validate that our proposed model is superior to the counterparts. In addition, the best SNPs identified by our model have been replicated in the literature, which justifies our prediction.

AB - With the rapid development of high-throughput genotyping and neuroimaging techniques, imaging genetics has drawn significant attention in the study of complex brain diseases such as Alzheimer's disease (AD). Research on the associations between genotype and phenotype improves the understanding of the genetic basis and biological mechanisms of brain structure and function. AD is a progressive neurodegenerative disease; therefore, the study on the relationship between single nucleotide polymorphism (SNP) and longitudinal variations of neuroimaging phenotype is crucial. Although some machine learning models have recently been proposed to capture longitudinal patterns in genotype-phenotype association studies, most machine-learning models base the learning on fixed structure among longitudinal prediction tasks rather than automatically learning the interrelationships. In response to this challenge, we propose a new automated time structure learning model to automatically reveal the longitudinal genotype-phenotype interactions and exploits such learned structure to enhance the phenotypic predictions. We proposed an efficient optimization algorithm for our model and provided rigorous theoretical convergence proof. We performed experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort for longitudinal phenotype prediction, including 3123 SNPs and 2 biomarkers (Voxel-Based Morphometry and FreeSurfer). The empirical results validate that our proposed model is superior to the counterparts. In addition, the best SNPs identified by our model have been replicated in the literature, which justifies our prediction.

KW - Alzheimer's disease

KW - genotype-phenotype association prediction

KW - longitudinal study

KW - low-rank model

KW - temporal structure auto-learning

UR - http://www.scopus.com/inward/record.url?scp=85050286574&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85050286574&partnerID=8YFLogxK

U2 - 10.1089/cmb.2018.0008

DO - 10.1089/cmb.2018.0008

M3 - Article

C2 - 30011249

AN - SCOPUS:85050286574

VL - 25

SP - 809

EP - 824

JO - Journal of Computational Biology

JF - Journal of Computational Biology

SN - 1066-5277

IS - 7

ER -