Whole genome sequence analysis of blood lipid levels in >66,000 individuals
Calcific aortic stenosis is one of the most common valvular heart diseases and has no efficacious medical therapy. In a multi-ancestry genome wide association study of 14,451 individuals with calcific aortic stenosis in the Million Veteran Program, Small et al discovered 6 novel genomic regions in the disease. Novel findings included variants near CEP85L, FTO, SLMAP, CELSR2, MECOM, and CDAN1 – together implicating complex pathobiology with contributions by lipid metabolism, inflammation, cellular senescence, and adiposity.
For monogenic traits, dominance is largely straightforward: How a trait manifests (its phenotype) is governed by how many copies of a dominant and/or recessive allele one inherits. The picture for polygenic traits, driven by loci throughout the genome, is less clear: In what way do individual variants influence phenotype? To explore this question, Ben Neale and colleagues developed an approach called dominance LD score regression (d-LDSC) and applied it to UK Biobank data on 1,060 traits. They found that dosage-dependent additive effects accounted for the vast majority of heritable trait variation in this population.
Protein-truncating variants (PTVs) disrupt a gene’s protein product. To find associations between rare PTVs in two protein-coding genes — APOB and PCSK9 — and low-density lipoprotein (LDL) cholesterol levels and coronary heart disease (CHD) risk, CGM PI, Amit Khera, and colleagues analyzed genomes from participants from five NHLBI studies and exomes from the UK Biobank, totaling over 200,000 individuals. They identified PTVs in 0.4 percent of individuals and found that PTV carriers had both lower LDL cholesterol levels over time and a 49 percent reduction in CHD risk.
Blood Pressure Control Targets and Risk of Cardiovascular and Cerebrovascular Events After Intracerebral Hemorrhage
Intracerebral hemorrhage (ICH) survivors are at high risk for recurrent stroke and cardiovascular events. Blood pressure (BP) control represents the most potent intervention to lower these risks, but optimal treatment targets in this patient population remain unknown. In this manuscript by CGM Investigators Jonathan Rosand, Christopher Anderson, Alessandro Biffi and colleagues, more intensive BP control than current guideline recommendations had a significantly greater effect on reducing the risk of major adverse cardiovascular and cerebrovascular events and mortality in the months to years following the initial stroke. This work has important implications for the way blood pressure is managed following this devastating form of stroke, and warrants study in a dedicated, randomized controlled trial.
Validation of a predictive model for obstructive sleep apnea in people with Down syndrome
Detecting obstructive sleep apnea (OSA) is important to both prevent significant comorbidities in people with Down syndrome (DS) and untangle contributions to other behavioral and mental health diagnoses. However, laboratory-based polysomnograms are often poorly tolerated, unavailable, or not covered by health insurance for this population. This work published by CGM investigator Brian Skotko and colleagues leveraged a previously developed prediction model that held promise in identifying which people with DS might not have significant apnea. In a novel set of participants with DS, a clinically reliable screening tool for OSA in people with DS that bypasses the need for laboratory-based polysomnography (sleep studies) was not achieved. This work, importantly, indicates that patients with DS should continue to be monitored for OSA according to current healthcare guidelines.
Polygenic Scores Help Reduce Racial Disparities in Predictive Accuracy of Automated Type 1 Diabetes Classification Algorithms
Automated algorithms to identify individuals with type 1 diabetes using electronic health records are increasingly used in biomedical research. It is not known whether the accuracy of these algorithms differs by self-reported race. This manuscript by CGM investigators Miriam Udler, Jose Florez, and CGM associate member Alisa Manning and colleauges investigates whether polygenic scores improve identification of individuals with type 1 diabetes. Using two large hospital-based biobanks (Mass General Brigham [MGB] and BioMe) the group analyzed an established automated algorithm for identifying type 1 diabetes and compared it to two published polygenic scores for type 1 diabetes. Importantly, the automated algorithm was more likely to incorrectly assign a diagnosis of type 1 diabetes in self-reported non-White individuals than in self-reported White individuals. After incorporating polygenic scores into the MGB Biobank, the positive predictive value of the type 1 diabetes algorithm increased from 70 to 97% for self-reported White individuals (meaning that 97% of those predicted to have type 1 diabetes indeed had type 1 diabetes) and from 53 to 100% for self-reported non-White individuals. Similar results were found in BioMe. This work importantly illuminates the inherent problems with automated phenotyping algorithms, and the risks of exacerbating health disparities because of an increased risk of misclassification of individuals from underrepresented populations. Polygenic scores may be used to improve the performance of phenotyping algorithms and potentially reduce this disparity.
Precise cut-and-paste DNA insertion using engineered type V-K CRISPR-associated transposases
Genome editing technologies capable of generating large sequence insertions would obviate the need to develop custom patient-specific approaches, enabling the treatment of larger swaths of patients with diverse mutations using a single therapeutic. Towards this goal, CGM Investigators led by Ben Kleinstiver recently developed several engineered versions of CRISPR-associated transposases (CASTs) with improved properties that can insert large kilobase-scale DNA cargos into genomes. We engineered CAST enzymes that have dramatically improved safety by reducing their off-target genome-wide integrations, that have enhanced insertion purity and efficiency, and that function for the first time in human cells, positioning CASTs as a leading technology for kilobase-scale genome edits for a new class of genetic medicines.
Seven technologies to watch in 2023: CRISPR anywhere
Nature picks the top seven tools and techniques that they feel are positioned to have the greatest scientific impact in 2023, among them being the CRSPR-Cas9 work being done in CGM, Ben Kleinstiver‘s lab. This is an important honor not only for his lab, but his technician leading the project, Russell Walton.
People with the same body-mass index (BMI) can have different distributions of body fat, which could affect heart and metabolic disease risk. To look for associations between fat distribution and disease risk, CGM PI, Amit Khera, and colleagues used deep learning models to analyze whole-body MRI images of more than 40,000 people from the UK Biobank and quantify fat volumes at three anatomical locations. Using these data, they found an association between deep belly fat and increased risk of type 2 diabetes and coronary artery disease in people with the same BMI, as well as a link between hip and thigh fat and reduced disease risk. The study shows how fat distribution can affect disease risk independent of BMI.
The Center for Genomic Medicine (CGM) comprises one of the largest and most vibrant hubs of genomic medicine research in the world. The CGM includes 46 faculty collaborating to define the ‘genomic medicine cycle’ – which envisages a genomics community seeking to advance research from basic genomics research to ultimately using genome information for diagnostics and targeted therapeutics. All of our investigators are faculty at Harvard Medical School, and many members of our community are also investigators at the Broad Institute of Harvard and MIT.