O'DONNELL-LURIA LAB

The O'Donnell-Luria Lab at Boston Children’s Hospital and in the Translational Genomics Group at the Broad Institute of MIT and Harvard spends every day pursuing the goal of finding the missing heritability for undiagnosed patients with rare disorders. By increasing our understanding of underlying patterns and mechanisms of genetic diseases we aim to improve diagnosis of these challenging cases. The team is currently working on analysis approaches to improve our understanding of the coding and noncoding genome and technology approaches (e.g., short and long read genome and RNA-sequencing, proteomics) to improve the discovery of variants and genes associated with rare disease. We also have projects exploring the role of unannotated open reading frames in disease, identify mechanisms underlying incomplete penetrance, and improve genetic prevalence estimation. We have a particular interest in disorders of chromatin machinery, in connection to the patients that I see through the EpiChroma Genetics Clinic. We contribute to efforts to improve variant classification and gene-disease curation, through ClinGen. And we engage with a number of consortia including GREGoR, gnomAD, NeuroDev project, and CAGI. Welcome to our lab website!

Incomplete penetrance, rare disease, genome sequencing, regulatory variants

The lack of penetrance estimates of putative pathogenic variants associated with genetic disease is a major challenge when counseling affected families, estimating disease prediction in currently healthy individuals, and prenatal screening. Understanding the extent and biological mechanisms of variable penetrance is crucial for variant interpretation and has the potential to dramatically improve disease risk prediction for a wide range of severe genetic diseases.

We are investigating the underlying biological mechanisms of incomplete penetrance by large-scale analysis of the world’s largest dataset: the Genome Aggregation Database (gnomAD). The gnomAD dataset is depleted of individuals with severe pediatric disorders, still, there are observations of individuals with early-onset, highly penetrant pathogenic variants for dominant disease in this dataset, making gnomAD a valuable resource for investigating penetrance mechanisms. By investigating this cohort our goal is to unravel some of the mechanisms underlying incomplete penetrance.

Collaborate with us

Do you have a fascinating case of incomplete penetrance and are interested in investigating the underlying mechanisms in collaboration with the O’Donnell-Luria Lab? Reach out to us at sgudmund@broadinstitute.org 

RNA-seq, Structural variation, mitochondrial variants, noncoding variants, constraint metrics

Although the adoption of exome sequencing (ES) in clinical practice improved the diagnostic sensitivity for rare diseases, many if not the majority of patients with putatively Mendelian diseases lack a molecular diagnosis. Whole genome sequencing (WGS) emerged as a logical extension of ES, but has not yet substantially improved the diagnostic yield in rare disease, at least in part due to the challenge of attributing causality for the rare structural and non-coding variants identified in WGS. Transcriptome sequencing (RNA-seq) is a potential tool to bridge this diagnostic challenge, by identifying rare splice alterations or allele-specific expression outliers that are not yet reliably predicted from primary genomic sequence with currently available methods.

Prior work in the Translational Genomics Group by Beryl Cummings (link) highlighted the power of RNA-seq in improving the diagnostic yield in Mendelian muscle diseases. Our goal is to develop best practices for implementing RNA-seq and WGS in rare disease discovery. 

Collaborate with us

If you have a patient or cohort with an unsolved Mendelian disease, and are interested in applying RNA-seq to gene discovery or variant resolution, please reach out to Vijay Ganesh at vganesh@broadinstitute.org

Novel gene discovery, population genetics, rare disease

Once thought impossible, genes that arise through non-copying mechanisms such as de novo evolution, long non-coding RNA that become coding, and overprinting have not only been shown to produce stable proteins, but have also been implicated in disease. However, they are not included in most gene annotation lists resulting in “unannotated genes”. We are investigating the prevalence of these unannotated genes in the general population using the Genome Aggregation Database (gnomAD). Additionally, we are interested in understanding if these genes may play a role in rare Mendelian diseases and aid in diagnosis.

Collaborate with us

If you are interested in contributing novel proteins or genes, please contact us at marten@broadinstitute.org. 

Rare disease prevalence, method development, variant curation, gnomAD

Prevalence of a condition is an important factor in decision-making by researchers and pharmaceutical companies to determine goals and resource allocation, yet the prevalence of the vast majority of rare diseases is unknown. Increased knowledge about the prevalence of a condition, both globally and in specific sub-populations (e.g. Ashkenazi Jewish, East Asian, etc) is important to advocacy groups and foundations that aim to extend outreach and support to build a network that includes as many affected individuals and families as possible. Many rare disease groups cite a lack of known disease prevalence as a barrier to engagement with pharmaceutical companies and as a challenge to outreach. A method for calculating prevalence for autosomal recessive disease has been developed in close collaboration with the Translational genomics group at the Broad Institute . This method leverages our expertise in variant curation along with the Genome Aggregation Database (gnomAD), the largest human reference database in the world that was developed at the Broad Institute (Karczewski et al., 2019)."

Collaborate with us

Do you have questions or want to know more, contact us about this project at freq-calc@broadinstitute.org