What does a Scientific Curator do?

In this series we explore some important roles in the life sciences, and how their work impacts patients and families.

In bioinformatics, the role of a scientific curator involves extracting and organising data from the literature, to help ensure that tools or databases are kept up to date with the most accurate information available. In this blog, Scientific Curator Achchuthan Shanmugasundram shares his experience working at Genomics England and what motivates him in his role.

My role at Genomics England

I work as a Scientific Curator in the Knowledge Management and Data Services team at Genomics England. My main responsibilities involve developing gene panels in our knowledgebase, PanelApp. This is an open access resource that provides gene-disease associations grouped into virtual gene panels. These panels are then used throughout the in the NHS Genomic Medicine Service (GMS) and the Generation Study.

Although my daily tasks vary depending on the priorities of the team, they can usually be grouped into 4 different categories:

1) Assessment of gene-disease association evidence

My primary role is to curate the evidence that links genes to rare conditions. The aim is to curate enough evidence so that genes can be included on panels for genomic tests used in the NHS. I review comments submitted to PanelApp by experts from our scientific and clinical communities, read peer-reviewed scientific articles, and search for curated information from other databases.

I then use this to rate the level of evidence for gene-disease associations using a traffic-light based classification system, where green means there is sufficient evidence for it to be used in a diagnostic pipeline. More details on this classification process can be found in a previous blog.
 

2) Data Integration

I also work on integrating high quality curated data from other resources. This involves developing streamlined processes for data integration, parsing data, mapping terminologies, and making them available via PanelApp.

An example of this is the integration of Developmental Disorders panel from Gene2Phenotype resource.

3) Managing data updates

Managing PanelApp data releases for the GMS takes up a big portion of my time at work. This involves preparing and finalising reports to send to the NHS, liaising with other teams that use the gene panel data and the NHS Genomics Unit, and drafting required documentation and coordinating testing.

4) Providing subject matter expertise

As a biologist and curator of PanelApp, I work closely with engineers from the Knowledge Management team to provide expertise on developing new features and functionalities for the resource.

My journey to scientific curation

I graduated with a degree in Biochemistry from Imperial College London, during which I decided to pursue a career in academia. I was successful in obtaining a funded PhD position at the University of Liverpool, with research focused on annotating genomes and curating metabolic pathways for microbes infecting humans and animals. This gave me the opportunity to get first-hand experience of genomics and curation. I also did a postdoc on genome annotation and analysis of a hibernating rodent called the alpine marmot.

After this, I decided to work as a biocurator for a genome database for eukaryotic pathogens, VEuPathDB. I chose curation because of my background in biology and ability to analyse complex information, which fit well with the curation role.

After working with VEuPathDB for 6 years, I decided to move to Genomics England. I knew that my work here could potentially have a positive impact on lives of patients, and I was drawn to the opportunity of hybrid working that Genomics England offers.

Although my previous experience provided me with the right skillset for this position, the challenge for me was to move into the world of rare conditions, step away from my academic mindset, and adapt to different ways of working. These include agile working, adhering to standards, and maintaining relationships with stakeholders in the genomics ecosystem.

The impact of my work on patient outcomes

The Rare Disease (RD) tiering pipeline sorts and prioritises potentially disease-causing genetic changes from the millions identified when someone has whole genome sequencing. This tiering process uses the virtual gene panels stored in PanelApp, which also hosts gene panels used for other genetic testing within the NHS. It serves as the reference resource for gene panel tests for Rare Disease and Germline Cancer Susceptibility listed in the National Genomic Test Directory.

The curation process adds the latest scientific knowledge to the gene panels, which could result in giving a diagnosis to a patient, potentially leading to a better patient outcome.

Furthermore, PanelApp is an open-source tool used by researchers, healthcare providers and diagnostic labs beyond the UK, potentially creating a positive impact for those affected by rare conditions across the globe.

What makes me passionate about my work?

Firstly, I am inspired by the stories of patients and families. I feel satisfied when I add new green genes to panels, as they can potentially lead to a diagnosis for people.

Secondly, I enjoy the varied nature of my job, particularly learning about rare conditions, and the management of delivering data releases.

Thirdly, it has been a great experience working in a multidisciplinary environment in Genomics England. I learn a lot through close interacts with people that have different skills and specialisms.

The future of curation at Genomics England

Curation is a manual and labour-intensive process, so the curation team have started to think about how we could increase the efficiency of our workflows. We have recently collaborated on a project with the Machine Learning team, where we used natural language processing to prioritise peer-reviewed publications for curation. We also plan to develop methods for automating different parts of the curation process.

There are many groups working on the genomics of rare conditions worldwide, and we are exploring how we can best use and share this knowledge for the benefit of all. There will be a lot of groundwork needed to assess the quality of data, map terminologies between resources, and develop strategies on how best to assign gene-disease associations to panels.

I hope to contribute to future developments in both these domains.