2019
DOI: 10.1038/s41467-019-09718-5
|View full text |Cite
|
Sign up to set email alerts
|

Abstract: Polygenic risk scores (PRS) have shown promise in predicting human complex traits and diseases. Here, we present PRS-CS, a polygenic prediction method that infers posterior effect sizes of single nucleotide polymorphisms (SNPs) using genome-wide association summary statistics and an external linkage disequilibrium (LD) reference panel. PRS-CS utilizes a high-dimensional Bayesian regression framework, and is distinct from previous work by placing a continuous shrinkage (CS) prior on SNP effect sizes, which is r… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

3
501
0
1

Year Published

2019
2019
2024
2024

Publication Types

Select...
7
3

Relationship

2
8

Authors

Journals

citations
Cited by 861 publications
(505 citation statements)
references
References 65 publications
3
501
0
1
Order By: Relevance
“…We conducted a simple GWAS without any covariates in group A. Next, we calculate weights for each of the common variants using PRS-CS (Ge et al 2019) , which takes the GWAS summary statistics from group A and an external linkage disequilibrium (LD) reference panel to estimate effect sizes in group B. Using these weights, we calculated PRS's for individuals in group C using PLINK (Chang et al 2015) with its built in allelic scoring procedure (--score).…”
Section: Computing Genomic Auc (Auc_prs)mentioning
confidence: 99%
“…We conducted a simple GWAS without any covariates in group A. Next, we calculate weights for each of the common variants using PRS-CS (Ge et al 2019) , which takes the GWAS summary statistics from group A and an external linkage disequilibrium (LD) reference panel to estimate effect sizes in group B. Using these weights, we calculated PRS's for individuals in group C using PLINK (Chang et al 2015) with its built in allelic scoring procedure (--score).…”
Section: Computing Genomic Auc (Auc_prs)mentioning
confidence: 99%
“…We retained SNPs with minor allele frequency > 0.01 and INFO quality score > 0.80 for scoring. To generate polygenic risk scores, we applied PRS-CS 23 other Bayesian methods, particularly when using a large discovery GWAS 23 (for comparison with conventional clumping and thresholding, see Methods S4). We set the global shrinkage parameter at 0.01 to reflect the likely polygenic architecture of major depression.…”
Section: Polygenic Risk Scoringmentioning
confidence: 99%
“…The simplest form of PRS construction requires two things: a selected set of independent risk variants with estimated or weighted risk effect sizes (say PRS construction methods and their underlying variant selection procedures can roughly be categorized into four groups: (i) fixed P-value thresholds of independent risk variants, e.g., "GWAS hits," variants that reached genome-wide significance (with ܲ ൏ 5 ൈ 1 0 ି ଼ ; (ii) LD pruning (actually clumping) / P-value thresholding ("P&T") of summary statistics that increases power by determining the most predictive P-value cutoff that can be above or below genome-significance [6]; (iii) genome-wide PRS that consider the full GWAS summary statistics after modeling LD, applying shrinkage or Bayesian approaches, e.g., LDpred and lassosum, [21][22][23][24] and (iv) methods that use individual-level data from a GWAS to determine an optimal set of independent predictors through Bayesian spike and slab or mixture priors [25]. The first two approaches typically use the originally reported effect sizes for weighting, while the latter two approaches model LD and/or shrink effect sizes.…”
Section: Introductionmentioning
confidence: 99%