As such, it

is helpful to observe similarities in heritability estimates across different methods of assessment. Multivariate analyses have explored the degree to which different PEs share genetic and environmental influences. Whether for individual PEs [10••], individual schizotypal domains [12], or symptom counts from different types of personality disorder [14], all studies reported considerable overlap in genetic effects across different PEs. For example, in a recent study of adolescents, paranoia and hallucinations correlated r = .47, and 64% of this covariation was explained by genetic influences, and the genetic correlation was high (0.61). Together, the multivariate results suggest considerable pleiotropic genetic effects across the different individual types of PE, together with some genetic effects being specific to individual PEs. Twin studies can also explore the ERK inhibitor degree to which causal influences on PEs are shared with other forms of psychopathology, cognition, and personality (for recent findings see 14, 16, 17, 18 and 19]). Table 2 outlines the two molecular genetic publications on PEs in general population samples on genome-wide identified variants. Overall, both studies, which employed adolescent samples, found some tentative evidence that genome-wide significant

variants associated with schizophrenia also influence variance in PEs in the community, as well as several negative results. One genome-wide significant schizophrenia-associated risk allele (rs17512836, in TCF4) was significantly associated Anacetrapib with higher INCB018424 quantitative scores on a paranoia scale in the general population at age 16 [20••]. TCF4 (transcription factor 4 gene) encodes a basic Helix-Loop-Helix (bHLH) transcription factor and is highly expressed in the brain, where it plays a role in neurodevelopment [21]. On the other hand, a second study, which used

a categorical score of presence of at least one definite PE at age 12 or 18, found no individual schizophrenia-associated variants to be significantly associated with their measure of PEs [22••]. Polygenic risk scores (the weighted sum of the number of risk alleles carried by an individual [23••]) were also employed in both studies in Table 2. Schizophrenia and bipolar disorder polygenic risk scores did not significantly predict any of six quantitative PE subscales at age 16 [20••] (scores were derived from the Psychiatric Genomics Consortium (PGC) stage-1 mega-analysis). The same schizophrenia polygenic risk score was investigated in the second study and did not predict the presence of at least one definite PE at either age 12 or 18 [22••]. Notably, individuals who had at least one definite PE had on average higher schizophrenia polygenic risk scores than those who had not had at least one PE [22••]. In sum, both studies provide some evidence for a genetic link between PEs in adolescence and diagnosed schizophrenia, but both studies also report negative findings.