Genomic Prediction of Complex Disease Risk

Abstract

We construct risk predictors using polygenic scores (PGS) computed from common Single Nucleotide Polymorphisms (SNPs) for a number of complex disease conditions, using L1-penalized regression (also known as LASSO) on case-control data from UK Biobank. Among the disease conditions studied are Hypothyroidism, (Resistant) Hypertension, Type 1 and 2 Diabetes, Breast Cancer, Prostate Cancer, Testicular Cancer, Gallstones, Glaucoma, Gout, Atrial Fibrillation, High Cholesterol, Asthma, Basal Cell Carcinoma, Malignant Melanoma, and Heart Attack. We obtain values for the area under the receiver operating characteristic curves (AUC) in the range ∼ 0.58 − 0.71 using SNP data alone. Substantially higher predictor AUCs are obtained when incorporating additional variables such as age and sex. Some SNP predictors alone are sufficient to identify outliers (e.g., in the 99th percentile of PGS) with 3 − 8 times higher risk than typical individuals. We validate predictors out-of-sample using the eMERGE dataset, and also with different ancestry subgroups within the UK Biobank population. Our results indicate that substantial improvements in predictive power are attainable using training sets with larger case populations. We anticipate rapid improvement in genomic prediction as more case-control data become available for analysis.

Human genotype-to-phenotype predictions: boosting accuracy with nonlinear models

“…eBMD also has a relatively high heritability [30,31] and there are efforts to predict it in [24] and [32] with up to r 2 = 0.25 and r 2 = 0.12. Asthma and hypothyroidism have a high prevalence (11% and 5% in the UK Biobank data) and are predicted in [4,25]. Psoriasis was selected to check how our methods work with a highly imbalanced phenotype (1% prevalence).…”

“…The genotype and phenotype data for our experiments were obtained from the full release of the UK Biobank [10]. We selected a cohort of 429,351 individuals of white British ethnicity according to data-field 21000 'Ethnic background' of the biobank so as to maintain a homogeneous population structure as in [4,5,24,25]. This cohort was further subdivided into train, validation, and test sets.…”

“…For example, Qian et al [4] using Snpnet, however they did not use any covariates other than age, sex, and principal components of the genotype matrix. Lello et al [25] report an AUC of 0.632, but on genetic data only and on a test set composed of self-reported white but non-genetically British individuals from the UK Biobank.…”

Genetic Risk Score for Predicting Schizophrenia Using Human Chromosomal-Scale Length Variation

“…Genetic risk scores [9][10][11] have been developed for many different forms of disease, including breast cancer [12], coronary artery disease [13], and stroke [14]. Polygenic risk scores based on SNPs clearly can predict schizophrenia.…”

“…Polygenic risk scores based on SNPs clearly can predict schizophrenia. One study measured an odds ratio of about 8 (95% CI [4][5][6][7][8][9][10][11][12][13][14] for the highest decile compared to the lowest decile [15]. A second study found that polygenic risk scores for schizophrenia (and bipolar disorder) are also associated with creativity [16].…”