Thread by @Joni_Coleman

Next is Omar Shanta who will talk about a GWAS of CNVs in 500k individuals (127k cases, mostly EUR, some AFR). Needed consistent CNV calling across entire cohort. Difficult platforms, which makes things challenging. #WCPG2022

Aims: what is the contribution of rare CNVs, which are those CNVs per disorder, how do rare CNVs compare across disorders? #WCPG2022

Assess multiple metrics of CNV burden, comparing tier 1 (pli > 0.5) and tier 2 (pLI < 0.5). #WCPG2022

CNVs in genes continue to risk, enticed for tier 1, particularly in ASD, ADHD and SCZ. Moderate effects elsewhere #WCPG2022

Combining different types of data EUR case-control, AFR case-control, EUR families. #WCPG2022

Gene level contributions seen across SCZ and ASD, including in known loci, and also see effects cross-disorder. At breakpoint level get extra loci, including some (ASTN2) that are only found across disorders #WCPG2022

How do CNVs compare across disorders? Breakpoint analysis identified novel loci. Pleiotropic loci as well, NRXN1, 1q21 #WCPG2022

Intersected data with brain expression showed risk loci differ by disorders and by CNV type (deletion increased in ASD, and duplication in SCZ). #WCPG2022

Next is @vivek12252 talking about cross disorder GWAS in iPSYCH. Initial iPSYCH cross disorder GWAS highlighted fetal development as a key pathway #WCPG2022

IPSYCH is a Danish national case-control cohort. IPSYCH 2015 doubled the 2012 sample size. See high genetic correlation between disorders in the waves of iPSYCH, larger than those from elsewhere e.g. the brainstorm consortium #WCPG2022

Genetic correlations in iPSYCH are stable and unique - much higher between waves in iPSYCH than between iPSYCH and PGC #WCPG2022

GenomicSEM supports a common factor model of disorders. Carried out a series of GWAS based on this - psych Vs control, case-case, Case-cohort GWAS #WCPG2022

Psych Vs control identified 5 new loci. Different designs emphasize different genetic components, as shown by lower correlations across designs compared to within designs #WCPG2022

Finally we have @AdrianAskelund who will present on the genomics of psychiatric symptom differentiation in early life. Mental health struggles in children tend to be considered as internalising (emotional) and externalising (behavioural) difficulties #WCPG2022

Generated differentiation scores as E - B and total scores as E+B. Analysis in MoBA - 41% of pregnant women from Norway in 1998. 240k[?] individuals, 80% genotyped #WCPG2022

This study: 79k children, 13k sibling pairs. Differentiation and total scores at 1.5-5 years and symptoms at 8 years. 11 PRS. #WCPG2022

Differentiation becomes more variable from 1.5 to 5. Differentiation predicts expected disorders at 8 (high --> conduct disorder etc., Low --> anxiety etc.) #WCPG2022

Assessing across disorder PRS - ADHD predicts tendency to behavioural differentiation, BIP as well to an extent. PTSD and SCZ predict tendency to emotional differentiation to an extent #WCPG2022

Next step is Genomic SEM #WCPG2022

Take homes: find support for effect of ADHD PRS on behavioural difficulties. May be helpful in refining prediction in child psychiatry #WCPG2022