Open Targets Genetics is well known as a powerful solution for both GWAS data processing, integration and visualization. However, what few people know is that – with the right knowledge – it is also highly configurable and can be extended to suit your specific needs.
Changing the pipeline
Before the data is displayed in Open Targets Genetics (OTG), it needs to go through multiple data processing pipelines. The input of these pipelines consists of GWAS data in combination with meta information describing the study. Each pipeline is responsible for a method that adds value to the GWAS studies. In general, these methods can be divided into variant extraction (fine-mapping), variant association and colocalization between studies. The results of each of these methods are loaded and visualized in OTG. The pipelines are well optimized for processing GWAS studies from specific sources such as the UK BioBank but that does not mean they cannot be changed.
There are three ways in which the pipeline can be changed; replacing existing tools and dependencies, adding additional tools and reconfiguring existing tools. In this blog post, we will cover the replacement and addition of tools.
Changing the fine-mapping pipeline
Fine-mapping is an important method used for analyzing GWAS studies as it extracts potentially associated variants (called the credible set). In the original pipeline, CGTA-COJO is used to perform the fine-mapping. However, many other tools are available that are based on different statistical methods to detect such credible sets. One of these tools is SuSiE. SuSiE is available as an R-package and selects variants based on the Sum of Single Effect, hence the name SuSiE. Without going into too much detail, we have shown that the output of SuSiE can be converted into a valid OTG format. This way, either the CGTA-COJO can be replaced by the SuSiE credible set or the data can be loaded on top of the existing data.
For one of our projects, we have added SuSiE on top of the existing data and extended the user interface with an additional Manhattan plot. As shown in the figure below, the Manhattan plot that is displayed on the study page is used to visualise the SuSiE results. This way, both fine-mapping methods complement each other which could uncover potential leads to new drug targets otherwise not found.
Extending the pipeline with MAGMA
MAGMA is a well-known tool for gene set analysis on GWAS data. In the public Open Targets Genetics application, genes are associated with traits via multiple data sources and methods. MAGMA is not one of them, but of course, that does not mean there is no value in adding it. Luckily, the codebase of the Open Targets Genetics is very well structured and is basically organized into a gene, variant and study centric view. As MAGMA will associate genes with studies, we can add this information to the gene and study page. By reusing the components that are part of OTGP's user-friendly interface we can display the MAGMA data as shown below.
Adding additional features
There are many methods to enrich GWAS studies with additional information and we can imagine that everyone has their own preferences. With Open Targets Genetics as a flexible solution to store, analyse and visualize GWAS data, it is perfectly suited for your preferred GWAS analysis methods. At The Hyve we have both the domain knowledge and technical expertise to create a GWAS pipeline tailored to your needs.
Please contact us for more information.