Predicting risk for Alcohol Use Disorder using longitudinal data with multimodal biomarkers and family history: a machine learning study

Sivan Kinreich, Jacquelyn L. Meyers, Adi Maron-Katz, Chella Kamarajan, Ashwini K. Pandey, David B. Chorlian, Jian Zhang, Gayathri Pandey, Stacey Subbie-Saenz de Viteri, Dan Pitti, Andrey P. Anokhin, Lance Bauer, Victor Hesselbrock, Marc A. Schuckit, Howard J. Edenberg, Bernice Porjesz

Research output: Contribution to journalArticle

Abstract

Predictive models have succeeded in distinguishing between individuals with Alcohol use Disorder (AUD) and controls. However, predictive models identifying who is prone to develop AUD and the biomarkers indicating a predisposition to AUD are still unclear. Our sample (n = 656) included offspring and non-offspring of European American (EA) and African American (AA) ancestry from the Collaborative Study of the Genetics of Alcoholism (COGA) who were recruited as early as age 12 and were unaffected at first assessment and reassessed years later as AUD (DSM-5) (n = 328) or unaffected (n = 328). Machine learning analysis was performed for 220 EEG measures, 149 alcohol-related single nucleotide polymorphisms (SNPs) from a recent large Genome-wide Association Study (GWAS) of alcohol use/misuse and two family history (mother DSM-5 AUD and father DSM-5 AUD) features using supervised, Linear Support Vector Machine (SVM) classifier to test which features assessed before developing AUD predict those who go on to develop AUD. Age, gender, and ancestry stratified analyses were performed. Results indicate significant and higher accuracy rates for the AA compared with the EA prediction models and a higher model accuracy trend among females compared with males for both ancestries. Combined EEG and SNP features model outperformed models based on only EEG features or only SNP features for both EA and AA samples. This multidimensional superiority was confirmed in a follow-up analysis in the AA age groups (12–15, 16–19, 20–30) and EA age group (16–19). In both ancestry samples, the youngest age group achieved higher accuracy score than the two other older age groups. Maternal AUD increased the model’s accuracy in both ancestries’ samples. Several discriminative EEG measures and SNPs features were identified, including lower posterior gamma, higher slow wave connectivity (delta, theta, alpha), higher frontal gamma ratio, higher beta correlation in the parietal area, and 5 SNPs: rs4780836, rs2605140, rs11690265, rs692854, and rs13380649. Results highlight the significance of sampling uniformity followed by stratified (e.g., ancestry, gender, developmental period) analysis, and wider selection of features, to generate better prediction scores allowing a more accurate estimation of AUD development.

Original languageEnglish (US)
JournalMolecular Psychiatry
DOIs
StateAccepted/In press - Jan 1 2019

Fingerprint

Biomarkers
Alcohols
Single Nucleotide Polymorphism
African Americans
Electroencephalography
Age Groups
Machine Learning
Mothers
Genome-Wide Association Study
Fathers
Alcoholism
Demography

ASJC Scopus subject areas

  • Molecular Biology
  • Psychiatry and Mental health
  • Cellular and Molecular Neuroscience

Cite this

Kinreich, S., Meyers, J. L., Maron-Katz, A., Kamarajan, C., Pandey, A. K., Chorlian, D. B., ... Porjesz, B. (Accepted/In press). Predicting risk for Alcohol Use Disorder using longitudinal data with multimodal biomarkers and family history: a machine learning study. Molecular Psychiatry. https://doi.org/10.1038/s41380-019-0534-x

Predicting risk for Alcohol Use Disorder using longitudinal data with multimodal biomarkers and family history : a machine learning study. / Kinreich, Sivan; Meyers, Jacquelyn L.; Maron-Katz, Adi; Kamarajan, Chella; Pandey, Ashwini K.; Chorlian, David B.; Zhang, Jian; Pandey, Gayathri; Subbie-Saenz de Viteri, Stacey; Pitti, Dan; Anokhin, Andrey P.; Bauer, Lance; Hesselbrock, Victor; Schuckit, Marc A.; Edenberg, Howard J.; Porjesz, Bernice.

In: Molecular Psychiatry, 01.01.2019.

Research output: Contribution to journalArticle

Kinreich, S, Meyers, JL, Maron-Katz, A, Kamarajan, C, Pandey, AK, Chorlian, DB, Zhang, J, Pandey, G, Subbie-Saenz de Viteri, S, Pitti, D, Anokhin, AP, Bauer, L, Hesselbrock, V, Schuckit, MA, Edenberg, HJ & Porjesz, B 2019, 'Predicting risk for Alcohol Use Disorder using longitudinal data with multimodal biomarkers and family history: a machine learning study', Molecular Psychiatry. https://doi.org/10.1038/s41380-019-0534-x
Kinreich, Sivan ; Meyers, Jacquelyn L. ; Maron-Katz, Adi ; Kamarajan, Chella ; Pandey, Ashwini K. ; Chorlian, David B. ; Zhang, Jian ; Pandey, Gayathri ; Subbie-Saenz de Viteri, Stacey ; Pitti, Dan ; Anokhin, Andrey P. ; Bauer, Lance ; Hesselbrock, Victor ; Schuckit, Marc A. ; Edenberg, Howard J. ; Porjesz, Bernice. / Predicting risk for Alcohol Use Disorder using longitudinal data with multimodal biomarkers and family history : a machine learning study. In: Molecular Psychiatry. 2019.
@article{e1ede4d40ef746d187bea72770c1d6b3,
title = "Predicting risk for Alcohol Use Disorder using longitudinal data with multimodal biomarkers and family history: a machine learning study",
abstract = "Predictive models have succeeded in distinguishing between individuals with Alcohol use Disorder (AUD) and controls. However, predictive models identifying who is prone to develop AUD and the biomarkers indicating a predisposition to AUD are still unclear. Our sample (n = 656) included offspring and non-offspring of European American (EA) and African American (AA) ancestry from the Collaborative Study of the Genetics of Alcoholism (COGA) who were recruited as early as age 12 and were unaffected at first assessment and reassessed years later as AUD (DSM-5) (n = 328) or unaffected (n = 328). Machine learning analysis was performed for 220 EEG measures, 149 alcohol-related single nucleotide polymorphisms (SNPs) from a recent large Genome-wide Association Study (GWAS) of alcohol use/misuse and two family history (mother DSM-5 AUD and father DSM-5 AUD) features using supervised, Linear Support Vector Machine (SVM) classifier to test which features assessed before developing AUD predict those who go on to develop AUD. Age, gender, and ancestry stratified analyses were performed. Results indicate significant and higher accuracy rates for the AA compared with the EA prediction models and a higher model accuracy trend among females compared with males for both ancestries. Combined EEG and SNP features model outperformed models based on only EEG features or only SNP features for both EA and AA samples. This multidimensional superiority was confirmed in a follow-up analysis in the AA age groups (12–15, 16–19, 20–30) and EA age group (16–19). In both ancestry samples, the youngest age group achieved higher accuracy score than the two other older age groups. Maternal AUD increased the model’s accuracy in both ancestries’ samples. Several discriminative EEG measures and SNPs features were identified, including lower posterior gamma, higher slow wave connectivity (delta, theta, alpha), higher frontal gamma ratio, higher beta correlation in the parietal area, and 5 SNPs: rs4780836, rs2605140, rs11690265, rs692854, and rs13380649. Results highlight the significance of sampling uniformity followed by stratified (e.g., ancestry, gender, developmental period) analysis, and wider selection of features, to generate better prediction scores allowing a more accurate estimation of AUD development.",
author = "Sivan Kinreich and Meyers, {Jacquelyn L.} and Adi Maron-Katz and Chella Kamarajan and Pandey, {Ashwini K.} and Chorlian, {David B.} and Jian Zhang and Gayathri Pandey and {Subbie-Saenz de Viteri}, Stacey and Dan Pitti and Anokhin, {Andrey P.} and Lance Bauer and Victor Hesselbrock and Schuckit, {Marc A.} and Edenberg, {Howard J.} and Bernice Porjesz",
year = "2019",
month = "1",
day = "1",
doi = "10.1038/s41380-019-0534-x",
language = "English (US)",
journal = "Molecular Psychiatry",
issn = "1359-4184",
publisher = "Nature Publishing Group",

}

TY - JOUR

T1 - Predicting risk for Alcohol Use Disorder using longitudinal data with multimodal biomarkers and family history

T2 - a machine learning study

AU - Kinreich, Sivan

AU - Meyers, Jacquelyn L.

AU - Maron-Katz, Adi

AU - Kamarajan, Chella

AU - Pandey, Ashwini K.

AU - Chorlian, David B.

AU - Zhang, Jian

AU - Pandey, Gayathri

AU - Subbie-Saenz de Viteri, Stacey

AU - Pitti, Dan

AU - Anokhin, Andrey P.

AU - Bauer, Lance

AU - Hesselbrock, Victor

AU - Schuckit, Marc A.

AU - Edenberg, Howard J.

AU - Porjesz, Bernice

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Predictive models have succeeded in distinguishing between individuals with Alcohol use Disorder (AUD) and controls. However, predictive models identifying who is prone to develop AUD and the biomarkers indicating a predisposition to AUD are still unclear. Our sample (n = 656) included offspring and non-offspring of European American (EA) and African American (AA) ancestry from the Collaborative Study of the Genetics of Alcoholism (COGA) who were recruited as early as age 12 and were unaffected at first assessment and reassessed years later as AUD (DSM-5) (n = 328) or unaffected (n = 328). Machine learning analysis was performed for 220 EEG measures, 149 alcohol-related single nucleotide polymorphisms (SNPs) from a recent large Genome-wide Association Study (GWAS) of alcohol use/misuse and two family history (mother DSM-5 AUD and father DSM-5 AUD) features using supervised, Linear Support Vector Machine (SVM) classifier to test which features assessed before developing AUD predict those who go on to develop AUD. Age, gender, and ancestry stratified analyses were performed. Results indicate significant and higher accuracy rates for the AA compared with the EA prediction models and a higher model accuracy trend among females compared with males for both ancestries. Combined EEG and SNP features model outperformed models based on only EEG features or only SNP features for both EA and AA samples. This multidimensional superiority was confirmed in a follow-up analysis in the AA age groups (12–15, 16–19, 20–30) and EA age group (16–19). In both ancestry samples, the youngest age group achieved higher accuracy score than the two other older age groups. Maternal AUD increased the model’s accuracy in both ancestries’ samples. Several discriminative EEG measures and SNPs features were identified, including lower posterior gamma, higher slow wave connectivity (delta, theta, alpha), higher frontal gamma ratio, higher beta correlation in the parietal area, and 5 SNPs: rs4780836, rs2605140, rs11690265, rs692854, and rs13380649. Results highlight the significance of sampling uniformity followed by stratified (e.g., ancestry, gender, developmental period) analysis, and wider selection of features, to generate better prediction scores allowing a more accurate estimation of AUD development.

AB - Predictive models have succeeded in distinguishing between individuals with Alcohol use Disorder (AUD) and controls. However, predictive models identifying who is prone to develop AUD and the biomarkers indicating a predisposition to AUD are still unclear. Our sample (n = 656) included offspring and non-offspring of European American (EA) and African American (AA) ancestry from the Collaborative Study of the Genetics of Alcoholism (COGA) who were recruited as early as age 12 and were unaffected at first assessment and reassessed years later as AUD (DSM-5) (n = 328) or unaffected (n = 328). Machine learning analysis was performed for 220 EEG measures, 149 alcohol-related single nucleotide polymorphisms (SNPs) from a recent large Genome-wide Association Study (GWAS) of alcohol use/misuse and two family history (mother DSM-5 AUD and father DSM-5 AUD) features using supervised, Linear Support Vector Machine (SVM) classifier to test which features assessed before developing AUD predict those who go on to develop AUD. Age, gender, and ancestry stratified analyses were performed. Results indicate significant and higher accuracy rates for the AA compared with the EA prediction models and a higher model accuracy trend among females compared with males for both ancestries. Combined EEG and SNP features model outperformed models based on only EEG features or only SNP features for both EA and AA samples. This multidimensional superiority was confirmed in a follow-up analysis in the AA age groups (12–15, 16–19, 20–30) and EA age group (16–19). In both ancestry samples, the youngest age group achieved higher accuracy score than the two other older age groups. Maternal AUD increased the model’s accuracy in both ancestries’ samples. Several discriminative EEG measures and SNPs features were identified, including lower posterior gamma, higher slow wave connectivity (delta, theta, alpha), higher frontal gamma ratio, higher beta correlation in the parietal area, and 5 SNPs: rs4780836, rs2605140, rs11690265, rs692854, and rs13380649. Results highlight the significance of sampling uniformity followed by stratified (e.g., ancestry, gender, developmental period) analysis, and wider selection of features, to generate better prediction scores allowing a more accurate estimation of AUD development.

UR - http://www.scopus.com/inward/record.url?scp=85074569255&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074569255&partnerID=8YFLogxK

U2 - 10.1038/s41380-019-0534-x

DO - 10.1038/s41380-019-0534-x

M3 - Article

C2 - 31595034

AN - SCOPUS:85074569255

JO - Molecular Psychiatry

JF - Molecular Psychiatry

SN - 1359-4184

ER -