NEW YORK – While the common genetic variants contributing to schizophrenia in East Asian individuals appear to broadly resemble those found in Europeans, polygenic risk scores (PRS) for the condition cannot necessarily be transferred from one population to the next, according to a new analysis by investigators in Singapore, the US, China, and elsewhere.
The study, published online today in Nature Genetics, "highlights the importance of including all major ancestral groups in genomic studies, both as a strategy to improve the power to find disease associations and to ensure that the findings have maximum relevance for all populations," co-first authors Max Lam and Chia-Yen Chen, researchers affiliated with Massachusetts General Hospital, the Broad Institute, and other centers in the US or in Singapore, and their co-authors wrote.
Although many common variants and loci have been implicated in schizophrenia, the team noted that most past studies have focused on individuals of European ancestry. That suggests far more variants are left to be identified but also makes it difficult to know how well the known associations between schizophrenia and variants hold up in other populations.
"Importantly, the applicability of training data from [European] studies to those of [non-European] ancestry has not been fully assessed," the authors explained, "leaving uncertainty as to the biological relevance of discoveries made in [European] sample for non-Europeans."
For their comparative analysis, the researchers performed a genome-wide association study focused on 22,778 East Asian individuals with schizophrenia and 35,362 without, shoring up findings from the first 13,305 cases and 16,244 controls with additional validation and meta-analysis. Their analyses led to 19 loci with 21 genome-wide significant ties to schizophrenia in East Asians, including 14 associations not identified in a prior schizophrenia study in individuals with Chinese ancestry.
The risk loci included many parts of the genome where allele frequencies differ in East Asian and European populations, the team noted, although analyses that incorporated linkage disequilibrium and common variant clues suggested that the genetic architecture of schizophrenia risk is comparable for parts of the genome outside of the immune-related major histocompatibility complex region.
Likewise, the researchers saw similar selection signatures, overrepresented gene sets, and other similarities when they compared the East Asian genetic data with results from prior schizophrenia studies in Europeans.
With a meta-analysis on East Asian and European individuals — including 56,418 cases and more than 78,800 controls — they highlighted 208 schizophrenia associations at 123 known and 53 new loci, narrowing in on apparent causal variants at 44 loci with additional fine-mapping analyses.
Even so, PRSs for schizophrenia that were developed in Europeans performed more poorly in the East Asian individuals and vice versa — reduced predictive abilities that the authors suspected might stem from allele frequency, linkage disequilibrium, and other population-related differences.
"[A]lthough individual loci typically have the same direction and a similar magnitude across populations," they wrote, "aggregating variants that differentially tag causal loci across populations for genetic risk prediction results in considerable variability in prediction accuracy."