Answer: GWAS stands for a “genone-wide association study.” Genome-wide association studies are used by researchers to determine the relationship between certain genes and diseases or other health outcomes.
A typical Manhattan plot, useful for displaying data collected from a genome wide association study.
Genome-wide association studies (GWAS) are an analytic observational technique used by neuroscientists to examine the relationship between genes and various health outcomes. GWAS are not experimental, but instead are used to develop hypotheses regarding the nervous system. Because a GWAS searches for commonalities in genes, they are most likely to yield results when the disease or condition has a strong heritability component.
The human genome is made up of more than 3 billion base pairs, coded by the letters A, T, G, and C. Typically, humans share 99.9% of the same genome with each other, but there are minor differences in their genomes that could account for people’s likelihood to develop neurological diseases or psychiatric conditions. These single nucleotide differences at particular sections of the genome are interesting in neuroscience and biomedical research. These types of mutations are commonly called single nucleotide polymorphisms, or SNPs.
A GWAS is typically set up as a case-control study. Consider if the researcher had a hypothesis that there is a relationship between certain SNPs and tardive dyskinesia among people with schizophrenia. They would start by collecting samples of DNA from two populations, people with schizophrenia and tardive dyskinesia, while the control group for comparison might be people with schizophrenia WITHOUT tardive dyskinesia, or people without schizophrenia with tardive dyskinesia. Then, from the DNA, it is possible to examine the differences in their genomes to see which SNPs are more prevalent among the experimental group.
Often times, results from a GWAS are displayed in a Manhattan plot. Manhattan plots are scatter plots. They plot the genomic locus on the x-axis, usually with chromosome numbers ranging from 1 on the far left and 22 on the far right. On the y-axis, these graphs plot the negative logarithm of the association p-value. When a SNP is strongly correlated with the condition, the p-value will be lower, and thus will be represented by a higher point on the plot. This type of graph is called a Manhattan plot because of its resemblance to the Manhattan skyline: occasionally, some genetic loci will appear much taller than the others surrounding it, like skyscrapers over the other buildings of the city. These high points of the plot indicate which of the genes are more associated with the condition of interest.
The GWAS is helpful in identifying the loci that contribute to elevated risk among multifactorial conditions, those which are influenced by the action of several genes.
Use of GWAS in neuroscience
The strength of a genome-wide association study is that new experimental hypotheses can be developed based on the results of the study. For example, consider a GWAS that compares a group of people with Alzheimer’s disease compared to people without. The GWAS may reveal that there are particular genetic loci that are much more prevalent across the Alzheimer’s group compared to the control group. Then, the experimenters may look to see what proteins these differences are responsible for coding. In the case of Alzheimer's disease, SNPs have been found across several genetic loci, ranging from the well established apolipoprotein to some weaker associations such as GAB2 (GRB2-associated binding protein 2), GALP (galanin-like peptide), PGBD1 ( piggyBac transposable element derived and others (Genome-wide association studies in Alzheimers disease).
Finally, the researchers may conduct manipulation studies where and how those proteins are changed to see differences in the outcomes in an experimental (non-human) model for Alzheimer’s disease.
Consider a different neurodegenerative disease with a heritability component, Parkinson's disease. These GWAS studies have demonstrated a strong link between the genetic locus LRRK2, and weaker links at 17 other loci including BST1, SNCA, HLA-DRA, CCDC62/HIP1R and MAPT (GWAS risk factors in Parkinson’s disease: LRRK2 coding variation and genetic interaction with PARK16). From this finding, the next step would be for an experimental study in nonhuman models of PD, where these particular protein products could be mutated or knocked out to determine some effect on outcome.
The more complex the condition, the more results that may appear in a GWAS. Many studies are concerned with identifying genetic factors associated with the presentation of autism and autism spectrum disorders in people. Many of the genes are associated with developmental markers, including FOXP1 (forkhead family of transcription factors), ATP2B2 (ATPase plasma membrane Ca2+ transporting 2), EXT1 (exostosin glycosyltransferase 1), and a handful of others (Meta-analysis of GWAS of over 16,000 individuals with autism spectrum disorder highlights a novel locus at 10q24.32 and a significant overlap with schizophrenia)