Schizophrenia Disease Study

High-resolution patient stratification enables novel target discovery and rapidly identifies repurposing opportunities

View all

186

disease associated combinations of SNPs identified

5

different patient communities

85

genes found to be associated with schizophrenia patients

Overview

Schizophrenia is the most common psychotic disease, hypothesized to be caused by altered neurotransmission and neurochemical imbalances. Standard antipsychotic medications form the basis of most patients’ treatment regimens, but as many as 30% of patients are classified as treatment resistant.

We analyzed genotype data from 823 patients diagnosed with schizophrenia from the UK Biobank, and performed high-resolution patient stratification. Our combinatorial analytics generated 186 disease associated combinations of single nucleotide polymorphisms (SNPs), which clustered to form five different patient subgroups. These subgroups contain distinct combinations of genes that appear to have shared biological mechanisms, providing further insights into their impact on underlying disease pathology and patient phenotype.

We also identified 85 significant genes associated with schizophrenia patients. Although most of the highest-scoring genes have not previously been implicated in the disease, they have a strong mechanistic link to several pathways that are associated with schizophrenia pathogenesis, including altered neurotransmission, inflammation, and regulation of neuronal development and synaptic plasticity.

Mapping existing drug options for key genes identified as being significant in subgroups of schizophrenia patients, allows us to rapidly identify potential repurposing opportunities. This includes a gene that is targeted by a statin that has already been shown to reduce the negative symptoms of schizophrenia, providing validation of our methodology.

There are many different theories underpinning the pathophysiology of schizophrenia; most implicate dysregulation of neurotransmitters such as dopamine, serotonin, and glutamate, in addition to abnormalities in brain structure. A significant number of patients respond poorly to the current antipsychotic medications available, with approximately 30% classified as treatment resistant. This means that there is a need for the development of new, more effective treatments for schizophrenia, as well as a greater understanding of the underlying disease mechanisms that might be used as predictors of treatment response.

Existing GWAS studies have identified a large number of loci and genes associated with the disease. However, these have proved challenging to translate into novel therapies and useful prediction tools that demonstrate clinical benefit. This is largely due to the complexity of the disease, with the non-linear effects of interaction between multiple variants contributing to a patient’s phenotype.

Genotype data of 547,197 SNPs from 823 schizophrenia cases (ICD-10 code F20) and 1,706 gender-matched controls (1:2 case control ratio) were obtained. The control set was generated from randomly selected, gender-matched individuals who did not have any mental or behavioral disorders. The PrecisionLife platform took less than an hour to identify high-order combinatorial genomic signatures (up to five SNP genotypes in combination) in this dataset. The combinatorial signatures produced can be used to estimate a patient’s disease risk more accurately.

Mapping these disease-associated SNPs to the human reference genome allows us to identify clinically relevant target genes and explain disease mechanisms. In this study, we used the disease signatures discovered to identify drug repurposing candidates for highly significant disease-associated target.

Schizophrenia Disease Architecture

124

disesae risk loci (SNPs)

risk-associated genes mapped to disesae-associated mechanisms

druggable targets

155

tool compounds

We identified 85 significant genes associated with schizophrenia patients. Although most have not previously been implicated in the disease, they have a strong mechanistic link to several pathways that are associated with schizophrenia pathogenesis, including altered neurotransmission, inflammation, and regulation of neuronal development and synaptic plasticity. Clustering the SNP networks based on their co-occurrence in patients revealed distinct sets of genes with shared biological pathways, indicating differences in drivers of disease mechanisms between groups of patients.

The disease signatures clustered to identify five different patient sub-cohorts. These sub-groups contain distinct combinations of genes that appear to have shared biological mechanisms, providing further insights into their impact on underlying disease pathology and patient phenotype. This high-resolution patient stratification based on distinct biological mechanisms presents significant opportunities in improving diagnosis and treatment of schizophrenia.

In addition, mapping existing drug options for key targets identified as being significant in subgroups of schizophrenia patients allows us to rapidly identify potential repurposing opportunities. Our results included a gene that is targeted by a statin that has already been shown to reduce the negative symptoms of schizophrenia, providing validation for our methodology.