A Giant, Genomic Discovery

Finding molecular diagnoses for rare pediatric disorders is challenging – and something that the Deciphering Developmental Disorders (DDD) study has been dedicated to for over a decade. In partnership with the Wellcome Sanger Institute, the project provides world-class expertise in genomic sequencing and computational analysis for the NHS genetic services. Their latest publication revealed that – using their advanced genome-sequencing methods – 5,500 children were successfully diagnosed with rare genetic diseases. And this is only the beginning. Though recruitment ended in 2015, the study continues to 2030.

We spoke with Helen Firth, senior co-author and clinical lead of the DDD study, about ethical concerns, challenges, and future hopes.

What are the project’s main aims?
 

The project has two main aims – discovery and diagnosis. We wanted to discover the genomic architecture of severe developmental disorders. We also planned to transform the way that diagnosis of such disorders is made in the NHS by pioneering the use of genome-wide sequencing with a trio-exome approach.

The study was inspired by the poor diagnostic rate in clinical practice for children with severe developmental disorders, which – prior to the DDD study – was only around 25 percent. With the novel genome-wide approaches used in the study, this figure has risen to  around 40 percent. There is still a lot to learn and discover to find diagnoses for the remaining children.

How was recruitment conducted?
 

We decided to focus our study on babies and young children with severe disorders that affect their development. In terms of neurodevelopmental disorders and congenital anomalies, we chose this demographic because it has a high clinical impact. We recruited patients through the NHS genetic services and focused on patients that had high suspicion of a monogenic cause – despite negative genetic testing on the routine tests available in the NHS at that time. 

Each individual genome contains four to five million variants where the sequence differs from the reference sequence. Finding the causative variant (or variants) for a rare disorder among this wealth of benign variation was a huge challenge. We therefore chose disorders of infancy and early childhood – since they were likely to be caused by variants that, at least for monoallelic disorders, were unlikely to be found in a healthy adult population.

What types of genetic testing are used in the study, and how are results analyzed and interpreted?
 

We used a trio-exome approach supplemented by a high-resolution array with five probes per exon of every gene and a SNP array. Combining these three types of data has been helpful in developing algorithms to detect copy number variants (CNVs) in exome data.

What are some of the most important insights that have emerged from the DDD study so far?
 

The most important discovery is that around 75 percent of the diagnoses of severe developmental disorders in monoallelic genes are due to de novo variants. We’ve found diagnoses for around 5,500 children across the UK whose variants fall in one of 800 genes. For many of the diagnoses we have made, there is just a single child in the study – indicating the enormous heterogeneity of developmental disorders. The study has shown how combining data and expertise across the NHS genetic services is crucially important. And it has taught us that a large dataset is needed for discovery and to evaluate rare variants.

We’ve discovered 60 new genes for developmental disorders, and the project has been a key contributor to ~300 papers in the peer-reviewed literature. The DDD study has laid the foundations for the 100,000 genomes project and the NHS Genome Medicine Service.

Were there any ethical challenges for such a large study?
 

The DDD study raised several ethical issues; however, we included an ethicist in our management committee from the start and employed a genetic counselor/social scientist to complete studies of public attitudes to genomic testing. In our latest publication, we have included a table of the ethical challenges we faced and how we addressed them (1).

What other hurdles did you have to overcome?
 

The main challenge was coordinating such a large study across 24 participating centers in the UK and Ireland. Setting up the systems to recruit 13,500 families, sequence 33,500 exomes, and report candidate diagnostic variants back to the recruiting centers for evaluation was far from easy! We used the Human Phenotype Ontology (HPO) for phenotyping and worked with their team to develop and optimize the HPO so that it had the scope (>10,000 terms) to enable accurate capture of the diverse clinical features encountered in the study. We also systematically collected detailed morphometric and developmental milestone data and information about family history, pregnancy, and perinatal events. This rich phenotypic data has been invaluable in interpreting the genomic data.

What lies on the horizon?
 

DDD is facilitated by the DECIPHER informatics platform, which is based at the European Bioinformatics Center. DECIPHER correlates phenotype with rare variants to map the clinically relevant variants in the human genome. The informatics infrastructure used in the DDD study and the knowledge gained are shared via DECIPHER to facilitate future diagnosis for patients globally. 

Also looking to the future, DDD continues to search for diagnoses for the undiagnosed participants and, when diagnoses are found, we immediately report back to the patient’s clinical geneticist. We hope to continue making new discoveries and have been amazed at how successful this collaboration between patients, clinicians, scientists, bioinformaticians, ethicists, and patient support groups – such as Unique – has been in finding answers for patients and their families.