PT - JOURNAL ARTICLE AU - Bogdan Pasaniuc AU - Alkes L. Price TI - Dissecting the genetics of complex traits using summary association statistics AID - 10.1101/072934 DP - 2016 Jan 01 TA - bioRxiv PG - 072934 4099 - http://biorxiv.org/content/early/2016/09/01/072934.short 4100 - http://biorxiv.org/content/early/2016/09/01/072934.full AB - During the past decade, genome-wide association studies (GWAS) have successfully identified tens of thousands of genetic variants associated with complex traits and diseases. These studies have produced vast repositories of genetic variation and trait measurements across millions of individuals, providing tremendous opportunities for further analyses. However, privacy concerns and other logistical considerations often limit access to individual-level genetic data, motivating the development of methods that analyze summary association statistics. Here we review recent progress on statistical methods that leverage summary association data to gain insights into the genetic basis of complex traits and diseases.INDIVIDUAL-LEVEL DATAGenome-wide SNP genotypes and trait values for each individual included in a GWAS.SUMMARY ASSOCIATION STATISTICSEstimated effect sizes and their standard errors for each SNP analyzed in a GWAS.Z-SCORESAssociation statistics that follow a standard normal distribution under the null; often computed as per-allele effect sizes divided by their standard error.META-ANALYSISA method for combining data from different studies in which summary association statistics from each study are jointly analyzed.MEGA-ANALYSISA method for combining data from different studies in which individual-level data from each study are merged and jointly analyzed.SUMMARY LD INFORMATIONIn-sample correlations between each pair of typed SNPs analyzed in a GWAS; can be restricted to proximal pairs of typed SNPs to limit the number of pairs of SNPs.TRANSCRIPTOME-WIDE ASSOCIATION STUDY (TWAS)A study that evaluates the association between expression of each gene and a trait of interest; predicted expression may be used instead of measured expression to improve practicality.MENDELIAN RANDOMIZATIONA method that uses significantly associated SNPs as instrumental variables to quantify causal relationships between two traits.BURDEN TESTA gene-based rare variant test in which all rare variants in a gene are assumed to have the same direction of effect.OVERDISPERSION TESTA gene-based rare variant test in which rare variants in a gene are assumed to impact trait in either direction.POSTERIOR PROBABILITY OF CAUSALITYThe inferred probability that a SNP is causal, based on association data and optional prior information.POLYGENIC RISK SCOREA method of predicting trait by summing the predicted marginal effects of all markers below a P-value threshold in a training sample, multiplied by marker genotypes in a validation sample.LD SCORE REGRESSIONA method of assessing trait polygenicity by regressing χ2 association statistics against LD scores for each SNP, computed as sums of squared correlations of each SNP with all SNPs including itself.PLEIOTROPYThe existence of shared genetic variant(s) with nonzero causal effect sizes for two traits.GENETIC CORRELATIONThe signed correlation across SNPs between causal effect sizes for two traits.