선택한 단어 수는 입니다.
최소 단어 이상 선택하여야 합니다.
최소 단어 이상 선택하여야 합니다.
선택한 단어 수는 30입니다.
최대 10 단어까지만 선택 가능합니다.
최대 10 단어까지만 선택 가능합니다.
문제 정의
- In this study, we investigated the effect of variable selection on the performance of prediction methods. Especially, we considered the following methods for variable selection and prediction: stepwise logistic regression (SLR), LASSO, and EN.
제안 방법
- In this study, we compared the performance of risk prediction models combining variable selection and prediction methods. Also, the effect of five SNP sets (i.e., ALL, ASIAN, KARE, ALL ＋ KARE, and ASIAN ＋ KARE) on risk prediction performance was investigated. Overall, we confirmed that prediction models incorporating both demographic variables and genetic variables were more accurate than prediction models using only demographic variables.
- For the joint identification of disease susceptibility variants among a large number of SNPs, we extracted SNPs having a strong correlation with T2D via logistic regression for single-variant analysis and collected the list of reported SNPs from a GWAS catalog [34]. Then, we implemented a 3-stage procedure as follows: the first stage was variable selection using SLR, LASSO, and EN.
- The external replication dataset combined two cohorts: the Health2 study (1,816 individuals) and Health Examinee (HEXA) study (3,696 individuals). In both the internal and external validation datasets, we measured the discriminative accuracy of the prediction models using AUC.
- In this study, we compared the performance of risk prediction models combining variable selection and prediction methods. Also, the effect of five SNP sets (i.
- First, we collected the SNPs, p1, from a GWAS catalog in all populations and an Asian population only. Second, the SNPs were selected by single-variant association test using logistic regression, with adjustments for age, sex, area (namely, rural area of Anseong and urban area of Ansan), and body mass index (BMI). We chose the top-ranked p2 SNPs by the order of p-values from the KARE cohort.
- The association of T2D was analyzed using logistic regression with adjustments for age, sex, area, and BMI as covariates. As shown in Supplementary Fig.
- Traditional approaches for disease risk prediction have been based on gene scores (GSs) [3-6]. The marginal effects of previously known disease-associated loci are estimated, and then, their sum can be used to construct a risk prediction model. While GS-based approaches can be useful when a genetic variant is responsible for diseases [7], they show low prediction performance when multiple genetic variants exist for a complex disease [8, 9].
- We compared the effect of variable selection on the performance of prediction by applying them to T2D GWAS chip data. We constructed the prediction models by combining variable selection and prediction methods using the Korean Association Resource (KARE) GWAS dataset (3,180 individuals) and then evaluated the performance of the risk prediction model through both internal validation (805 individuals in the KARE testing dataset) and external validation (4,723 individuals in an external replication dataset). The external replication dataset combined two cohorts: the Health2 study (1,816 individuals) and Health Examinee (HEXA) study (3,696 individuals).
대상 데이터
- Our investigation was based on the analysis of an external replication dataset of 4,723 samples (1,112 T2D subjects, 3,611 normal subjects) for T2D [33]. The criteria for grouping T2D subjects (nHealth2 = 794, nHEXA = 318) and nondiabetic normal subjects (nHealth2 = 770, nHEXA = 2,841) were FPG level (FPG ≥ 126 for T2D subjects and FPG ≤ 100 for nondiabetic normal subjects) and history of T2D treatment.
이론/모형
- λ and α are a tuning parameter and a weight of a value between 0 to 1, respectively. All penalized methods were conducted using the R-package glmnet [37].
- For the given covariates, the selection of SNPs was determined by a stepwise procedure based on Akaike's information criterion [35].
- For the given covariates, the selection of SNPs was determined by a stepwise procedure based on Akaike's information criterion [35]. The stepwise procedure was conducted using the R-package MASS [36].
- 1). The variable selection was performed using 5-fold cross-validation (CV) on the training set. We describe the details below.
성능/효과
- Overall, we confirmed that prediction models incorporating both demographic variables and genetic variables were more accurate than prediction models using only demographic variables. According to our results, the best combinations were SLR-LASSO and SLR-EN with three set from the KARE SNP set in the KARE testing dataset, whereas the SLR-SLR and SLR-EN combination with one set from the ALL ＋ KARE SNP set outperformed all other combinations in an external replication dataset.
- 816 in the KARE testing dataset. In an external replication dataset, SLR-SLR and SLR-EN with one set from ALL ＋ KARE (AUC, 0.726) were the best, with 51 SNPs for T2D, while SLR-LASSO and SLR-EN with three set from KARE (AUC, 0.590) showed the best performance, with 53 SNPs. SLR-SLR with one set from ALL ＋ KARE was superior to the model with only demographic variables (15.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.