Polygenic risk scores have shown great promise in predicting complex disease risk and will become more accurate as training sample sizes increase. The standard approach for calculating risk scores involves linkage disequilibrium (LD)-based marker pruning and applying a p value threshold to association statistics, but this discards information and can reduce predictive accuracy. We introduce LDpred, a method that infers the posterior mean effect size of each marker by using a prior on effect sizes and LD information from an external reference panel. Theory and simulations show that LDpred outperforms the approach of pruning followed by thresholding, particularly at large sample sizes. Accordingly, predicted R(2) increased from 20.1% to 25.3% in a large schizophrenia dataset and from 9.8% to 12.0% in a large multiple sclerosis dataset. A similar relative improvement in accuracy was observed for three additional large disease datasets and for non-European schizophrenia samples. The advantage of LDpred over existing methods will grow as sample sizes increase.

Copy number variants (CNVs) have been strongly implicated in the genetic etiology of schizophrenia (SCZ). However, genome-wide investigation of the contribution of CNV to risk has been hampered by limited sample sizes. We sought to address this obstacle by applying a centralized analysis pipeline to a SCZ cohort of 21,094 cases and 20,227 controls. A global enrichment of CNV burden was observed in cases (OR=1.11, P=5.7×10−15), which persisted after excluding loci implicated in previous studies (OR=1.07, P=1.7 ×10−6). CNV burden was enriched for genes associated with synaptic function (OR = 1.68, P = 2.8 ×10−11) and neurobehavioral phenotypes in mouse (OR = 1.18, P= 7.3 ×10−5). Genome-wide significant evidence was obtained for eight loci, including 1q21.1, 2p16.3 (NRXN1), 3q29, 7q11.2, 15q13.3, distal 16p11.2, proximal 16p11.2 and 22q11.2. Suggestive support was found for eight additional candidate susceptibility and protective loci, which consisted predominantly of CNVs mediated by non-allelic homologous recombination.

We carried out a genome-wide association study of schizophrenia (479 cases, 2,937 controls) and tested loci with P < 10(-5) in up to 16,726 additional subjects. Of 12 loci followed up, 3 had strong independent support (P < 5 x 10(-4)), and the overall pattern of replication was unlikely to occur by chance (P = 9 x 10(-8)). Meta-analysis provided strongest evidence for association around ZNF804A (P = 1.61 x 10(-7)) and this strengthened when the affected phenotype included bipolar disorder (P = 9.96 x 10(-9)).

Several lines of evidence have placed the catechol-O-methyltransferase (COMT) gene in the limelight as a candidate gene for schizophrenia. One of these is its biochemical function in metabolism of catecholamine neurotransmitters; another is the microdeletion, on chromosome 22q11, that includes the COMT gene and causes velocardiofacial syndrome, a syndrome associated with a high rate of psychosis, particularly schizophrenia. The interest in the COMT gene as a candidate risk factor for schizophrenia has led to numerous linkage and association analyses. These, however, have failed to produce any conclusive result. Here we report an efficient approach to gene discovery. The approach consists of (i) a large sample size-to our knowledge, the present study is the largest case-control study performed to date in schizophrenia; (ii) the use of Ashkenazi Jews, a well defined homogeneous population; and (iii) a stepwise procedure in which several single nucleotide polymorphisms (SNPs) are scanned in DNA pools, followed by individual genotyping and haplotype analysis of the relevant SNPs. We found a highly significant association between schizophrenia and a COMT haplotype (P=9.5x10-8). The approach presented can be widely implemented for the genetic dissection of other common diseases.

"Selective genotyping" is the term used when the determination of linkage between marker loci and quantitative trait loci (QTL) affecting some particular trait is carried out by genotyping only individuals from the high and low phenotypic tails of the entire sample population. Selective genotyping can markedly decrease the number of individuals genotyped for a given power at the expense of an increase in the number of individuals phenotyped. The optimum proportion of individuals genotyped from the point of view of minimizing costs for a given experimental power depends strongly on the cost of completely genotyping an individual for all of the markers included in the experiment (including the costs of obtaining a DNA sample) relative to the cost of rearing and trait evaluation of an individual. However, in single trait studies, it will almost never be useful to genotype more than the upper and lower 25% of a population. It is shown that the observed difference in quantitative trait values associated with alternative marker genotypes in the selected population can be much greater than the actual gene effect at the quantitative trait locus when the entire population is considered. An expression and a figure is provided for converting observed differences under selective genotyping to actual gene effects.