Tagging SNP-set selection with maximum information based on linkage disequilibrium structure in genome-wide association studies

Obesity risk in young adults from the Jerusalem Perinatal Study (JPS): the contribution of polygenic risk and early life exposure

BACKGROUND/OBJECTIVES

The effects of early life exposures on offspring life-course health are well established. This study assessed whether adding early socio-demographic and perinatal variables to a model based on polygenic risk score (PRS) improves prediction of obesity risk.

METHODS

We used the Jerusalem Perinatal study (JPS) with data at birth and body mass index (BMI) and waist circumference (WC) measured at age 32. The PRS was constructed using over 2.1M common SNPs identified in genome-wide association study (GWAS) for BMI. Linear and logistic models were applied in a stepwise approach. We first examined the associations between genetic variables and obesity-related phenotypes (e.g., BMI and WC). Secondly, socio-demographic variables were added and finally perinatal exposures, such as maternal pre-pregnancy BMI (mppBMI) and gestational weight gain (GWG) were added to the model. Improvement in prediction of each step was assessed using measures of model discrimination (area under the curve, AUC), net reclassification improvement (NRI) and integrated discrimination improvement (IDI).

RESULTS

One standard deviation (SD) change in PRS was associated with a significant increase in BMI (β = 1.40) and WC (β = 2.45). These associations were slightly attenuated (13.7-14.2%) with the addition of early life exposures to the model. Also, higher mppBMI was associated with increased offspring BMI (β = 0.39) and WC (β = 0.79) (p < 0.001). For obesity (BMI ≥ 30) prediction, the addition of early socio-demographic and perinatal exposures to the PRS model significantly increased AUC from 0.69 to 0.73. At an obesity risk threshold of 15%, the addition of early socio-demographic and perinatal exposures to the PRS model provided a significant improvement in reclassification of obesity (NRI, 0.147; 95% CI 0.068-0.225).

CONCLUSIONS

Inclusion of early life exposures, such as mppBMI and maternal smoking, to a model based on PRS improves obesity risk prediction in an Israeli population-sample.

The body mass index increases the genetic risk scores' ability to predict risk of hepatic damage in European adolescents: The HELENA study

BACKGROUND

Hepatic disorders are often complex and multifactorial, modulated by genetic and environmental determinants. During the last years, the hepatic disease has been progressively established from early stages in life. The use of genetic risk scores (GRS) to predict the genetic susceptibility to a particular phenotype among youth has gained interest in recent years. Moreover, the alanine aminotransferase (ALT) blood biomarker is often considered as hepatic screening tool, in combination with imaging techniques. The aim of the present study was to develop an ALT-specific GRS to help in the evaluation of hepatic damage risk in European adolescents.

METHODS

A total of 972 adolescents (51.3% females), aged 12.5-17.5 years, from the Healthy Lifestyle in Europe by Nutrition in Adolescence study were included in the analyses. The sample incorporated adolescents in all body mass index (BMI) categories and was divided considering healthy/unhealthy ALT levels, using sex-specific cut-off points. From 1212 a priori ALT-related single nucleotide polymorphisms (SNPs) extracted from candidate gene selection, a first screening of 234 SNPs univariately associated was established, selecting seven significant SNPs (p < .05) in the multivariate model. An unweighted GRS (uGRS) was developed by summing the number of reference alleles, and a weighted GRS (wGRS), by multiplying each allele to its estimated coefficient.

RESULTS

The uGRS and wGRS were significantly associated with ALT (p < .001). The area under curve was obtained integrating BMI as clinical factor, improving the predictive ability for uGRS (.7039) and wGRS (.7035), using 10-fold internal cross-validation.

CONCLUSIONS

Considering BMI status, both GRSs could contribute as complementary tools to help in the early diagnosis of hepatic damage risk in European adolescents.

Obesity Prediction in Young Adults from the Jerusalem Perinatal Study: Contribution of Polygenic Risk and Early Life Exposures

We assessed whether adding early life exposures to a model based on polygenic risk score (PRS) improves prediction of obesity risk. We used a birth cohort with data at birth and BMI and waist circumference (WC) measured at age 32. The PRS was composed of SNPs identified in GWAS for BMI. Linear and logistic models were used to explore associations with obesity-related phenotypes. Improvement in prediction was assessed using measures of model discrimination (AUC), and net reclassification improvement (NRI). One SD change in PRS was associated with a significant increase in BMI and WC. These associations were slightly attenuated (13.7%-14.2%) with the addition of early life exposures to the model. Also, higher maternal pre-pregnancy BMI was associated with increase in offspring BMI and WC (p<0.001). For prediction obesity (BMI ≥ 30), the addition of early life exposures to the PRS model significantly increase the AUC from 0.69 to 0.73. At an obesity risk threshold of 15%, the addition of early life exposures to the PRS model provided a significant improvement in reclassification of obesity (NRI, 0.147; 95% CI 0.068-0.225). We conclude that inclusion of early life exposures to a model based on PRS improves obesity risk prediction in an Israeli population-sample.

Development of a genetic risk score to predict the risk of hypertension in European adolescents from the HELENA study

Introduction

From genome wide association study (GWAS) a large number of single nucleotide polymorphisms (SNPs) have previously been associated with blood pressure (BP) levels. A combination of SNPs, forming a genetic risk score (GRS) could be considered as a useful genetic tool to identify individuals at risk of developing hypertension from early stages in life. Therefore, the aim of our study was to build a GRS being able to predict the genetic predisposition to hypertension (HTN) in European adolescents.

Methods

Data were extracted from the Healthy Lifestyle in Europe by Nutrition in Adolescence (HELENA) cross-sectional study. A total of 869 adolescents (53% female), aged 12.5-17.5, with complete genetic and BP information were included. The sample was divided into altered (≥130 mmHg for systolic and/or ≥80 mmHg for diastolic) or normal BP. Based on the literature, a total of 1.534 SNPs from 57 candidate genes related with BP were selected from the HELENA GWAS database.

Results

From 1,534 SNPs available, An initial screening of SNPs univariately associated with HTN (p < 0.10) was established, to finally obtain a number of 16 SNPs significantly associated with HTN (p < 0.05) in the multivariate model. The unweighted GRS (uGRS) and weighted GRS (wGRS) were estimated. To validate the GRSs, the area under the curve (AUC) was explored using ten-fold internal cross-validation for uGRS (0.802) and wGRS (0.777). Further covariates of interest were added to the analyses, obtaining a higher predictive ability (AUC values of uGRS: 0.879; wGRS: 0.881 for BMI z-score). Furthermore, the differences between AUCs obtained with and without the addition of covariates were statistically significant (p < 0.05).

Conclusions

Both GRSs, the uGRS and wGRS, could be useful to evaluate the predisposition to hypertension in European adolescents.

Prioritized candidate causal haplotype blocks in plant genome-wide association studies

Genome wide association studies (GWAS) can play an essential role in understanding genetic basis of complex traits in plants and animals. Conventional SNP-based linear mixed models (LMM) that marginally test single nucleotide polymorphisms (SNPs) have successfully identified many loci with major and minor effects in many GWAS. In plant, the relatively small population size in GWAS and the high genetic diversity found in many plant species can impede mapping efforts on complex traits. Here we present a novel haplotype-based trait fine-mapping framework, HapFM, to supplement current GWAS methods. HapFM uses genotype data to partition the genome into haplotype blocks, identifies haplotype clusters within each block, and then performs genome-wide haplotype fine-mapping to prioritize the candidate causal haplotype blocks of trait. We benchmarked HapFM, GEMMA, BSLMM, GMMAT, and BLINK in both simulated and real plant GWAS datasets. HapFM consistently resulted in higher mapping power than the other GWAS methods in high polygenicity simulation setting. Moreover, it resulted in smaller mapping intervals, especially in regions of high LD, achieved by prioritizing small candidate causal blocks in the larger haplotype blocks. In the Arabidopsis flowering time (FT10) datasets, HapFM identified four novel loci compared to GEMMA’s results, and the average mapping interval of HapFM was 9.6 times smaller than that of GEMMA. In conclusion, HapFM is tailored for plant GWAS to result in high mapping power on complex traits and improved on mapping resolution to facilitate crop improvement.

Genome-wide association studies (GWAS) are commonly used in human and plant studies to identify genetic variants responsible for the phenotype of interest and provide foundations for studying disease mechanisms and crop improvement. Most GWAS models are developed and optimized using human datasets. However, the difference between human and plant datasets essentially limits their applications in plant studies, especially when mapping complex traits such as drought resistance and yield. In this study, we present a novel GWAS method, HapFM, tailored for plant datasets to overcome the difficulties of many conventional GWAS methods. HapFM resulted in higher statistical power than conventional GWAS methods for mapping complex traits in our simulation and real dataset analyses. In addition, HapFM reduced the mapping interval by prioritizing candidate causal regions in the genome, which benefits the downstream experimental studies. Last but not least, HapFM can incorporate biological annotations to increase statistical power further. Overall, HapFM balances statistical power, result interpretability, and downstream experimental verifiability.

Correlation analysis of four KRTAP gene polymorphisms and cashmere fiber diameters in two cashmere goat breeds

Fiber diameter, a quantitative trait, is controlled by minor effect polygenes. Keratin-associated proteins (KRTAPs) are an important part of hair, and their rich polymorphisms facilitate the mining of cashmere trait molecular markers. In this study, Jiangnan and Tibetan cashmere goats were taken as the research object; multiplex PCR and exome sequencing technology were used to identify the exon regional polymorphisms of cashmere goats KRTAP15-1, KRTAP13.1, KRTAP27-1, and KRTAP24-1. The effects of mutation sites on the fiber diameter of cashmere were analyzed by least square method. The results showed that there were 28 mutation sites in the four KRTAP genes in Jiangnan cashmere goats and Tibetan cashmere goat populations. Among them, the KRTAP13.1, KRTAP27-1, and KRTAP24-1 gene polymorphisms were found to be significantly related to the fiber diameter of Jiangnan cashmere goats. The exploration of molecular markers in this study will help to improve the fiber diameter of the down, while the identification of gene polymorphisms will provide original data for the utilization and protection of germplasm resources of cashmere goats.

A new insight into the SNP genotyping using high-resolution melting method after the correlation analysis of the SNPs with WSSV-resistant traits

Procambarus clarkii is an important freshwater cultured crayfish in China. With the gradual development of its aquaculture industry, research on white spot disease, which is harmful to healthy culture of P. clarkii, increases gradually. The prophenoloxidase (proPO) system is an important part of crayfish's innate immunity and plays a role in virus resistance. In this study, based on the early discovery of three SNP sites in the intron of proPO gene, the linkage disequilibrium and haplotype were analyzed for the SNPs, and it was found that there was a strong linkage disequilibrium relationship among them. Through the analysis on association between the haplotypes and genotype of each SNP site with the WSSV-resistant traits, the detection of the SNP_7081 genotype was considered as the most convenient and efficient way for WSSV-resistant group selection. Furtherly, the high-resolution melting curve (HRM), which is a rapid and economic genotyping method, was chosen to establish for SNP_7081 site genotyping. The 68 bp target fragment with 27.94% GC content was amplified and melting curve analysis were performed. However, the appearance of false negatives which led to unable automatically grouped although the melting curves of genotypes CC, C>T and T>C were obviously different, and could be treated as standard to manually genotype the samples with an accuracy rate of 97.61%. The low GC content which correlated with the Tm value, was confirmed as the reason for the false negatives by the assay about the recombinant plasmid PMD18-T-SNP_7081 constructed with 45.24% GC content. Eventually, the adaptor primers were used to increase the GC content of the target fragment, and a modified HRM method for genotyping SNP_7081 site that could group automatically was established, which could provide a new insight for the HRM method to genotype SNPs.

Grouping of genomic markers in populations with family structure

BACKGROUND

Linkage and linkage disequilibrium (LD) between genome regions cause dependencies among genomic markers. Due to family stratification in populations with non-random mating in livestock or crop, the standard measures of population LD such as [Formula: see text] may be biased. Grouping of markers according to their interdependence needs to account for the actual population structure in order to allow proper inference in genome-based evaluations.

RESULTS

Given a matrix reflecting the strength of association between markers, groups are built successively using a greedy algorithm; largest groups are built at first. As an option, a representative marker is selected for each group. We provide an implementation of the grouping approach as a new function to the R package hscovar. This package enables the calculation of the theoretical covariance between biallelic markers for half- or full-sib families and the derivation of representative markers. In case studies, we have shown that the number of groups comprising dependent markers was smaller and representative SNPs were spread more uniformly over the investigated chromosome region when the family stratification was respected compared to a population-LD approach. In a simulation study, we observed that sensitivity and specificity of a genome-based association study improved if selection of representative markers took family structure into account.

CONCLUSIONS

Chromosome segments which frequently recombine in the underlying population can be identified from the matrix of pairwise dependence between markers. Representative markers can be exploited, for instance, for dimension reduction prior to a genome-based association study or the grouping structure itself can be employed in a grouped penalization approach.

Development of a Genetic Risk Score to predict the risk of overweight and obesity in European adolescents from the HELENA study

Obesity is the result of interactions between genes and environmental factors. Since monogenic etiology is only known in some obesity-related genes, a genetic risk score (GRS) could be useful to determine the genetic predisposition to obesity. Therefore, the aim of our study was to build a GRS able to predict genetic predisposition to overweight and obesity in European adolescents. A total of 1069 adolescents (51.3% female), aged 11-19 years participating in the Healthy Lifestyle in Europe by Nutrition in Adolescence (HELENA) cross-sectional study were genotyped. The sample was divided in non-overweight (non-OW) and overweight/obesity (OW/OB). From 611 single nucleotide polymorphisms (SNP) available, a first screening of 104 SNPs univariately associated with obesity (p < 0.20) was established selecting 21 significant SNPs (p < 0.05) in the multivariate model. Unweighted GRS (uGRS) was calculated by summing the number of risk alleles and weighted GRS (wGRS) by multiplying the risk alleles to each estimated coefficient. The area under curve (AUC) was calculated in uGRS (0.723) and wGRS (0.734) using tenfold internal cross-validation. Both uGRS and wGRS were significantly associated with body mass index (BMI) (p < .001). Both GRSs could potentially be considered as useful genetic tools to evaluate individual's predisposition to overweight/obesity in European adolescents.

Study of the Relationship between Polymorphisms in the IL-8 Gene Promoter Region and Coccidiosis Resistance Index in Jinghai Yellow Chickens

Interleukin 8 (IL-8) participates in the immune response and has the function of inducing neutrophils to release lysosomal enzymes and eliminate pathogens. This study was to investigate the effect of single nucleotide mutations in the IL-8 gene promoter region on the coccidiosis resistance index. In this study, 180 infected Eimeria tenella (E. tenella) Jinghai yellow chickens were used as experimental samples. DNA sequencing technology was used to detect single nucleotide polymorphisms (SNPs) in the IL-8 gene promoter region. The association between these SNPs and coccidiosis resistance indexes (including superoxide dismutase (SOD), malondialdehyde (MDA), glutathione peroxidase (GSH-PX), catalase (CAT), nitric oxide (NO), interleukin-1β (IL-1β), interleukin-2 (IL-2), interleukin-6 (IL-6), IL-8, and interferon-γ (IFN-γ)) were analyzed. Three SNPs (T-550C, G-398T, and T-360C) were detected. Significant associations were found between each genotype at the T-550C site with NO (p-value = 0.006) and IL-8 (p-value = 0.034) indexes. Significant associations were found between each genotype at the G-398T site with SOD (p-value = 0.042), CAT (p-value = 0.049), NO (p-value = 0.008), and IL-2 (p-value = 0.044) indexes. Significant associations were found between each genotype at the T-360C site with SOD (p-value = 0.007), NO (p-value = 0.046), IL-2 (p-value = 0.041), IL-8 (p-value = 0.039), and IFN-γ (p-value = 0.042) indexes. Haplotype analysis showed that multiple indexes of the H1H3 haplotype combination were significantly higher than other haplotype combinations. Therefore, mutation of the IL-8 gene promoter region has a significant regulatory effect on the coccidiosis resistance index, with a change in transcription factor binding potentially altering IL-8 gene expression, thereby further affecting the IL-8 level in plasma. However, the specific mechanism needs further study.

Predicting SSRI-Resistance: Clinical Features and tagSNPs Prediction Models Based on Support Vector Machine

BACKGROUND

A large proportion of major depressive patients will experience recurring episodes. Many patients still do not response to available antidepressants. In order to meaningfully predict who will not respond to which antidepressant, it may be necessary to combine multiple biomarkers and clinical variables.

METHODS

Eight hundred fifty-seven patients with recurrent major depressive disorder who were followed up 3-10 years involved 32 variables including socio-demographic, clinical features, and SSRIs treatment features when they received the first treatment. Also, 34 tagSNPs related to 5-HT signaling pathway, were detected by using mass spectrometry analysis. The training samples which had 12 clinical variables and four tagSNPs with statistical differences were learned repeatedly to establish prediction models based on support vector machine (SVM).

RESULTS

Twelve clinical features (psychomotor retardation, psychotic symptoms, suicidality, weight loss, SSRIs average dose, first-course treatment response, sleep disturbance, residual symptoms, personality, onset age, frequency of episode, and duration) were found significantly difference (P< 0.05) between 302 SSRI-resistance and 304 SSRI non-resistance group. Ten SSRI-resistance predicting models were finally selected by using support vector machine, and our study found that mutations in tagSNPs increased the accuracy of these models to a certain degree.

CONCLUSION

Using a data-driven machine learning method, we found 10 predictive models by mining existing clinical data, which might enable prospective identification of patients who are likely to resistance to SSRIs antidepressant.