Integrated genomic and BMI analysis for type 2 diabetes risk assessment

Dayanara Lebrón-Aldea, Emily J. Dhurandhar, Paulino Pérez-Rodríguez, Yann C Klimentidis, Hemant K. Tiwari, Ana I. Vazquez

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Type 2 Diabetes (T2D) is a chronic disease arising from the development of insulin absence or resistance within the body, and a complex interplay of environmental and genetic factors. The incidence of T2D has increased throughout the last few decades, together with the occurrence of the obesity epidemic. The consideration of variants identified by Genome Wide Association Studies (GWAS) into risk assessment models for T2D could aid in the identification of at-risk patients who could benefit from preventive medicine. In this study, we build several risk assessment models, evaluated with two different classification approaches (Logistic Regression and Neural Networks), to measure the effect of including genetic information in the prediction of T2D. We used data from to the Original and the Offspring cohorts of the Framingham Heart Study, which provides phenotypic and genetic information for 5,245 subjects (4,306 controls and 939 cases). Models were built by using several covariates: gender, exposure time, cohort, body mass index (BMI), and 65 SNPs associated to T2D. We fitted Logistic Regressions and Bayesian Regularized Neural Networks and then assessed their predictive ability by using a ten-fold crossvalidation. We found that the inclusion of genetic information into the risk assessment models increased the predictive ability by 2%, when compared to the baseline model. Furthermore, the models that included BMI at the onset of diabetes as a possible effector, gave an improvement of 6% in the area under the curve derived from the ROC analysis. The highest AUC achieved (0.75) belonged to the model that included BMI, and a genetic score based on the 65 established T2D associated SNPs. Finally, the inclusion of SNPs and BMI raised predictive ability in all models as expected; however, results from the AUC in Neural Networks and Logistic Regression did not differ significantly in their prediction accuracy.

Original languageEnglish (US)
Article number00075
JournalFrontiers in Genetics
Volume5
Issue numberFEB
DOIs
StatePublished - 2015

Fingerprint

Type 2 Diabetes Mellitus
Body Mass Index
Area Under Curve
Single Nucleotide Polymorphism
Logistic Models
Preventive Medicine
Genome-Wide Association Study
Inclusion Bodies
ROC Curve
Chronic Disease
Obesity
Insulin
Incidence

ASJC Scopus subject areas

  • Genetics
  • Molecular Medicine
  • Genetics(clinical)

Cite this

Lebrón-Aldea, D., Dhurandhar, E. J., Pérez-Rodríguez, P., Klimentidis, Y. C., Tiwari, H. K., & Vazquez, A. I. (2015). Integrated genomic and BMI analysis for type 2 diabetes risk assessment. Frontiers in Genetics, 5(FEB), [00075]. https://doi.org/10.3389/fgene.2015.00075

Integrated genomic and BMI analysis for type 2 diabetes risk assessment. / Lebrón-Aldea, Dayanara; Dhurandhar, Emily J.; Pérez-Rodríguez, Paulino; Klimentidis, Yann C; Tiwari, Hemant K.; Vazquez, Ana I.

In: Frontiers in Genetics, Vol. 5, No. FEB, 00075, 2015.

Research output: Contribution to journalArticle

Lebrón-Aldea, D, Dhurandhar, EJ, Pérez-Rodríguez, P, Klimentidis, YC, Tiwari, HK & Vazquez, AI 2015, 'Integrated genomic and BMI analysis for type 2 diabetes risk assessment', Frontiers in Genetics, vol. 5, no. FEB, 00075. https://doi.org/10.3389/fgene.2015.00075
Lebrón-Aldea D, Dhurandhar EJ, Pérez-Rodríguez P, Klimentidis YC, Tiwari HK, Vazquez AI. Integrated genomic and BMI analysis for type 2 diabetes risk assessment. Frontiers in Genetics. 2015;5(FEB). 00075. https://doi.org/10.3389/fgene.2015.00075
Lebrón-Aldea, Dayanara ; Dhurandhar, Emily J. ; Pérez-Rodríguez, Paulino ; Klimentidis, Yann C ; Tiwari, Hemant K. ; Vazquez, Ana I. / Integrated genomic and BMI analysis for type 2 diabetes risk assessment. In: Frontiers in Genetics. 2015 ; Vol. 5, No. FEB.
@article{548accf6b25e4ad0a81c5ca1af69dedf,
title = "Integrated genomic and BMI analysis for type 2 diabetes risk assessment",
abstract = "Type 2 Diabetes (T2D) is a chronic disease arising from the development of insulin absence or resistance within the body, and a complex interplay of environmental and genetic factors. The incidence of T2D has increased throughout the last few decades, together with the occurrence of the obesity epidemic. The consideration of variants identified by Genome Wide Association Studies (GWAS) into risk assessment models for T2D could aid in the identification of at-risk patients who could benefit from preventive medicine. In this study, we build several risk assessment models, evaluated with two different classification approaches (Logistic Regression and Neural Networks), to measure the effect of including genetic information in the prediction of T2D. We used data from to the Original and the Offspring cohorts of the Framingham Heart Study, which provides phenotypic and genetic information for 5,245 subjects (4,306 controls and 939 cases). Models were built by using several covariates: gender, exposure time, cohort, body mass index (BMI), and 65 SNPs associated to T2D. We fitted Logistic Regressions and Bayesian Regularized Neural Networks and then assessed their predictive ability by using a ten-fold crossvalidation. We found that the inclusion of genetic information into the risk assessment models increased the predictive ability by 2{\%}, when compared to the baseline model. Furthermore, the models that included BMI at the onset of diabetes as a possible effector, gave an improvement of 6{\%} in the area under the curve derived from the ROC analysis. The highest AUC achieved (0.75) belonged to the model that included BMI, and a genetic score based on the 65 established T2D associated SNPs. Finally, the inclusion of SNPs and BMI raised predictive ability in all models as expected; however, results from the AUC in Neural Networks and Logistic Regression did not differ significantly in their prediction accuracy.",
author = "Dayanara Lebr{\'o}n-Aldea and Dhurandhar, {Emily J.} and Paulino P{\'e}rez-Rodr{\'i}guez and Klimentidis, {Yann C} and Tiwari, {Hemant K.} and Vazquez, {Ana I.}",
year = "2015",
doi = "10.3389/fgene.2015.00075",
language = "English (US)",
volume = "5",
journal = "Frontiers in Genetics",
issn = "1664-8021",
publisher = "Frontiers Media S. A.",
number = "FEB",

}

TY - JOUR

T1 - Integrated genomic and BMI analysis for type 2 diabetes risk assessment

AU - Lebrón-Aldea, Dayanara

AU - Dhurandhar, Emily J.

AU - Pérez-Rodríguez, Paulino

AU - Klimentidis, Yann C

AU - Tiwari, Hemant K.

AU - Vazquez, Ana I.

PY - 2015

Y1 - 2015

N2 - Type 2 Diabetes (T2D) is a chronic disease arising from the development of insulin absence or resistance within the body, and a complex interplay of environmental and genetic factors. The incidence of T2D has increased throughout the last few decades, together with the occurrence of the obesity epidemic. The consideration of variants identified by Genome Wide Association Studies (GWAS) into risk assessment models for T2D could aid in the identification of at-risk patients who could benefit from preventive medicine. In this study, we build several risk assessment models, evaluated with two different classification approaches (Logistic Regression and Neural Networks), to measure the effect of including genetic information in the prediction of T2D. We used data from to the Original and the Offspring cohorts of the Framingham Heart Study, which provides phenotypic and genetic information for 5,245 subjects (4,306 controls and 939 cases). Models were built by using several covariates: gender, exposure time, cohort, body mass index (BMI), and 65 SNPs associated to T2D. We fitted Logistic Regressions and Bayesian Regularized Neural Networks and then assessed their predictive ability by using a ten-fold crossvalidation. We found that the inclusion of genetic information into the risk assessment models increased the predictive ability by 2%, when compared to the baseline model. Furthermore, the models that included BMI at the onset of diabetes as a possible effector, gave an improvement of 6% in the area under the curve derived from the ROC analysis. The highest AUC achieved (0.75) belonged to the model that included BMI, and a genetic score based on the 65 established T2D associated SNPs. Finally, the inclusion of SNPs and BMI raised predictive ability in all models as expected; however, results from the AUC in Neural Networks and Logistic Regression did not differ significantly in their prediction accuracy.

AB - Type 2 Diabetes (T2D) is a chronic disease arising from the development of insulin absence or resistance within the body, and a complex interplay of environmental and genetic factors. The incidence of T2D has increased throughout the last few decades, together with the occurrence of the obesity epidemic. The consideration of variants identified by Genome Wide Association Studies (GWAS) into risk assessment models for T2D could aid in the identification of at-risk patients who could benefit from preventive medicine. In this study, we build several risk assessment models, evaluated with two different classification approaches (Logistic Regression and Neural Networks), to measure the effect of including genetic information in the prediction of T2D. We used data from to the Original and the Offspring cohorts of the Framingham Heart Study, which provides phenotypic and genetic information for 5,245 subjects (4,306 controls and 939 cases). Models were built by using several covariates: gender, exposure time, cohort, body mass index (BMI), and 65 SNPs associated to T2D. We fitted Logistic Regressions and Bayesian Regularized Neural Networks and then assessed their predictive ability by using a ten-fold crossvalidation. We found that the inclusion of genetic information into the risk assessment models increased the predictive ability by 2%, when compared to the baseline model. Furthermore, the models that included BMI at the onset of diabetes as a possible effector, gave an improvement of 6% in the area under the curve derived from the ROC analysis. The highest AUC achieved (0.75) belonged to the model that included BMI, and a genetic score based on the 65 established T2D associated SNPs. Finally, the inclusion of SNPs and BMI raised predictive ability in all models as expected; however, results from the AUC in Neural Networks and Logistic Regression did not differ significantly in their prediction accuracy.

UR - http://www.scopus.com/inward/record.url?scp=84923551571&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84923551571&partnerID=8YFLogxK

U2 - 10.3389/fgene.2015.00075

DO - 10.3389/fgene.2015.00075

M3 - Article

VL - 5

JO - Frontiers in Genetics

JF - Frontiers in Genetics

SN - 1664-8021

IS - FEB

M1 - 00075

ER -