Stuart A Cook

Imperial College London, Londinium, England, United Kingdom

Are you Stuart A Cook?

Claim your profile

Publications (73)741.85 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Mutations in the gene encoding the RNA-binding protein RBM20 have been implicated in dilated cardiomyopathy (DCM), a major cause of chronic heart failure, presumably through altering cardiac RNA splicing. Here, we combined transcriptome-wide crosslinking immunoprecipitation (CLIP-seq), RNA-seq, and quantitative proteomics in cell culture and rat and human hearts to examine how RBM20 regulates alternative splicing in the heart. Our analyses revealed the presence of a distinct RBM20 RNA-recognition element that is predominantly found within intronic binding sites and linked to repression of exon splicing with RBM20 binding near 3' and 5' splice sites. Proteomic analysis determined that RBM20 interacts with both U1 and U2 small nuclear ribonucleic particles (snRNPs) and suggested that RBM20-dependent splicing repression occurs through spliceosome stalling at complex A. Direct RBM20 targets included several genes previously shown to be involved in DCM as well as genes not typically associated with this disease. In failing human hearts, reduced expression of RBM20 affected alternative splicing of several direct targets, indicating that differences in RBM20 expression may affect cardiac function. Together, these findings identify RBM20-regulated targets and provide insight into the pathogenesis of human heart failure.
    The Journal of clinical investigation. 06/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: Myocardial fibrosis identified by late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) in patients with hypertrophic cardiomyopathy (HCM) is associated with adverse cardiovascular events, but its value as an independent risk factor for sudden cardiac death (SCD) is unknown. We investigated the role of LGE-CMR in the risk stratification of HCM.
    Heart (British Cardiac Society) 06/2014; · 5.01 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The common heritable condition of Bicuspid Aortic Valve (BAV) is phenotypically heterogeneous, with valve dysfunction and aortopathy the major complications. We report overrepresentation of rare, likely pathogenic variants in target genes in a large cohort of 176 patients with BAV. We also describe a more severe aortic phenotype in patients with more than one known or likely pathogenic variant, supporting a multi-hit hypothesis for development of complications of BAV.
    Heart (British Cardiac Society) 06/2014; 100(Suppl 3):A55-A56. · 5.01 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Histone modifications are epigenetic marks that play fundamental roles in many biological processes including the control of chromatin-mediated regulation of gene expression. Little is known about interindividual variability of histone modification levels across the genome and to what extent they are influenced by genetic variation. We annotated the rat genome with histone modification maps, identified differences in histone trimethyl-lysine levels among strains, and described their underlying genetic basis at the genome-wide scale using ChIP-seq in heart and liver tissues in a panel of rat recombinant inbred and their progenitor strains. We identified extensive variation of histone methylation levels among individuals and mapped hundreds of underlying cis- and trans-acting loci throughout the genome that regulate histone methylation levels in an allele-specific manner. Interestingly, most histone methylation level variation was trans-linked and the most prominent QTL identified influenced H3K4me3 levels at 899 putative promoters throughout the genome in the heart. Cis- acting variation was enriched in binding sites of distinct transcription factors in heart and liver. The integrated analysis of DNA variation together with histone methylation and gene expression levels showed that histoneQTLs are an important predictor of gene expression and that a joint analysis significantly enhanced the prediction of gene expression traits (eQTLs). Our data suggest that genetic variation has a widespread impact on histone trimethylation marks that may help to uncover novel genotype-phenotype relationships.
    Genome Research 05/2014; · 14.40 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Alternative splicing is the main mechanism governing protein diversity. The recent developments in RNA-Seq technology have enabled the study of the global impact and regulation of this biological process. However, the lack of standardized protocols constitutes a major bottleneck in the analysis of alternative splicing. This is particularly important for the identification of exon-exon junctions, which is a critical step in any analysis workflow. Here we performed a systematic benchmarking of alignment tools to dissect the impact of design and method on the mapping, detection and quantification of splice junctions from multi-exon reads. Accordingly, we devised a novel pipeline based on TopHat2 combined with a splice junction detection algorithm, which we have named FineSplice. FineSplice allows effective elimination of spurious junction hits arising from artefactual alignments, achieving up to 99% precision in both real and simulated data sets and yielding superior F1 scores under most tested conditions. The proposed strategy conjugates an efficient mapping solution with a semi-supervised anomaly detection scheme to filter out false positives and allows reliable estimation of expressed junctions from the alignment output. Ultimately this provides more accurate information to identify meaningful splicing patterns. FineSplice is freely available at
    Nucleic Acids Research 02/2014; · 8.81 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Cardiac phenotypes, such as left ventricular (LV) mass,demonstrate high heritability although most genes associated with these complex traits remain unidentified. Genome-wide association studies (GWAS) have relied on conventional 2D cardiovascular magnetic resonance (CMR) as the gold-standard for phenotyping. However this technique is insensitive to the regional variations in wall thickness which are often associated with left ventricular hypertrophy and require large cohorts to reach significance. Here we test whether automated cardiac phenotyping using high spatial resolution CMR atlases can achieve improved precision for mapping wall thickness in healthy populations and whether smaller sample sizes are required compared to conventional methods. LV short-axis cine images were acquired in 138 healthy volunteers using standard 2D imaging and 3D high spatial resolution CMR. A multi-atlas technique was used to segment and co-register each image. The agreement between methods for end-diastolic volume and mass was made using Bland-Altman analysis in 20 subjects. The 3D and 2D segmentations of the LV were compared to manual labeling by the proportion of concordant voxels (Dice coefficient) and the distances separating corresponding points. Parametric and nonparametric data were analysed with paired t-tests and Wilcoxon signed-rank test respectively. Voxelwise power calculations used the interstudy variances of wall thickness. The 3D volumetric measurements showed no bias compared to 2D imaging. The segmented 3D images were more accurate than 2D images for defining the epicardium (Dice: 0.95vs0.93, P < 0.001; mean error 1.3mm vs 2.2mm, P < 0.001) and endocardium (Dice 0.95 vs 0.93, P < 0.001; mean error 1.1mm vs 2.0mm, P < 0.001).The 3D technique resulted in significant differences in wall thickness assessment at the base, septum and apex of the LV compared to 2D (P < 0.001). Fewer subjects were required for 3D imaging to detect a 1mm difference in wall thickness (72 vs 56, P < 0.001). High spatial resolution CMR with automated phenotyping provides greater power for mapping wall thickness than conventional 2D imaging and enables a reduction in the sample size required for studies of environmental and genetic determinants of LV wall thickness.
    Journal of Cardiovascular Magnetic Resonance 02/2014; 16(1):16. · 4.44 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: NECTAR (Non-synonymous Enriched Coding muTation ARchive; is a database and web application to annotate disease-related and functionally important amino acids in human proteins. A number of tools are available to facilitate the interpretation of DNA variants identified in diagnostic or research sequencing. These typically identify previous reports of DNA variation at a given genomic location, predict its effects on transcript and protein sequence and may predict downstream functional consequences. Previous reports and functional annotations are typically linked by the genomic location of the variant observed. NECTAR collates disease-causing variants and functionally important amino acid residues from a number of sources. Importantly, rather than simply linking annotations by a shared genomic location, NECTAR annotates variants of interest with details of previously reported variation affecting the same codon. This provides a much richer data set for the interpretation of a novel DNA variant. NECTAR also identifies functionally equivalent amino acid residues in evolutionarily related proteins (paralogues) and, where appropriate, transfers annotations between them. As well as accessing these data through a web interface, users can upload batches of variants in variant call format (VCF) for annotation on-the-fly. The database is freely available to download from the ftp site:
    Nucleic Acids Research 12/2013; · 8.81 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The study of families with rare inherited forms of hypo- and hypertension has been one of the most successful strategies to probe the molecular pathophysiology of blood pressure control and has revealed dysregulation of distal nephron sodium reabsorption to be a common mechanism. Familial Hyperkalaemic Hypertension (FHHt, Gordon Syndrome) is a salt-dependent form of hypertension caused by mutations in regulators of the thiazide-sensitive NaCl cotransporter, NCC, and is effectively treated by thiazide diuretics and/or dietary salt restriction. Variation in at least four genes can cause FHHt, including 'With No lysine (K)' kinases (WNK1 and WNK4), KeLcH-Like3 (KLHL3) and CULlin3 (CUL3). Here we identify novel disease-causing variants in CUL3 and KLHL3 segregating in 63% of pedigrees with previously unexplained FHHt, confirming the importance of these recently-described FHHt genes. We demonstrate conclusively, in two unrelated affected individuals, that rare intronic variants in CUL3 cause skipping of exon 9, as has been proposed. KLHL3 variants all occur in kelch-repeat domains and so are likely to disrupt WNK-complex binding. We found no evidence of plausible disease-causing variants within SLC4A8 (an alternative thiazide-sensitive sodium transporter) in this population. This supports existing evidence that CUL3 and KLHL3 gene products are physiologically important regulators of thiazide-sensitive distal nephron NaCl reabsorption, and hence potentially interesting novel anti-hypertensive drug targets. As a third of our non-WNK FHHt families do not have plausible CUL3 or KLHL3 variants, there are likely to be additional, as yet undiscovered, regulators of thiazide-sensitive pathways.
    Clinical Science 11/2013; · 4.86 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Distinguishing genetic variants that cause disease from variants that are rare but benign is one of the principal challenges in contemporary clinical genetics, particularly as variants are identified at a pace exceeding the capacity of researchers to characterise them functionally. We previously developed a novel method, called paralogue annotation, which accurately and specifically identifies disease-causing missense variants by transferring disease-causing annotations across families of related proteins. Here we refine our approach, and apply it to novel variants found in 2266 patients across two large cohorts with inherited sudden death syndromes, namely catecholaminergic polymorphic ventricular tachycardia (CPVT) or Brugada syndrome (BrS). Over one third of the novel non-synonymous variants found in these studies, which would otherwise be reported in a clinical diagnostics setting as 'variants of unknown significance', are categorised by our method as likely disease causing (positive predictive value 98.7%). This identified more than 500 new disease loci for BrS and CPVT. Our methodology is widely transferable across all human disease genes, with an estimated 150 000 potentially informative annotations in more than 1800 genes. We have developed a web resource that allows researchers and clinicians to annotate variants found in individuals with inherited arrhythmias, comprising a referenced compendium of known missense variants in these genes together with a user-friendly implementation of our approach. This tool will facilitate the interpretation of many novel variants that might otherwise remain unclassified.
    Journal of Medical Genetics 10/2013; · 5.70 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Cardiovascular Magnetic Resonance (CMR) is the gold-standard technique for assessment of ventricular function. Although left ventricular (LV) volumes and ejection fraction are strong predictors of outcome in dilated cardiomyopathy (DCM), there are limited data regarding the prognostic significance of right ventricular (RV) systolic dysfunction (RVSD). We investigated whether CMR assessment of RV function has prognostic value in DCM. We prospectively studied 250 consecutive DCM patients using CMR. RVSD, defined by RV ejection fraction ≤45%, was present in 86 (34%) patients. During a median follow-up period of 6.8 years, there were 52 deaths and 7 patients underwent cardiac transplantation . The primary end point of all-cause mortality or cardiac transplantation was reached by 42 of 86 patients with RVSD and 17 of 164 patients without RVSD (49% vs. 10%; hazard ratio [HR], 5.90; 95% confidence interval [CI], 3.35 to 10.37; P<0.001). On multivariable analysis, RVSD remained a significant independent predictor of the primary end point (HR, 3.90; 95% CI, 2.16 to 7.04; P<0.001), as well as secondary outcomes of cardiovascular mortality or cardiac transplantation (HR, 3.35; 95% CI, 1.76 to 6.39; P<0.001), and heart failure (HF) death, HF hospitalization or cardiac transplantation (HR, 2.70; 95% CI, 1.32 to 5.51; P=0.006). Assessment of RVSD improved risk stratification for all-cause mortality or cardiac transplantation (net reclassification improvement, 0.31; 95% CI 0.10 to 0.53; P=0.001). RVSD is a powerful, independent predictor of transplant-free survival and adverse HF outcomes in DCM. CMR assessment of RV function is important in the evaluation and risk stratification of DCM patients.
    Circulation 08/2013; · 15.20 Impact Factor
  • BMJ (online) 07/2013; 347:f3940. · 17.22 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Deletion 1p36 syndrome is recognized as the most common terminal deletion syndrome. Here, we describe the loss of a gene within the deletion that is responsible for the cardiomyopathy associated with monosomy 1p36, and we confirm its role in nonsyndromic left ventricular noncompaction cardiomyopathy (LVNC) and dilated cardiomyopathy (DCM). With our own data and publically available data from array comparative genomic hybridization (aCGH), we identified a minimal deletion for the cardiomyopathy associated with 1p36del syndrome that included only the terminal 14 exons of the transcription factor PRDM16 (PR domain containing 16), a gene that had previously been shown to direct brown fat determination and differentiation. Resequencing of PRDM16 in a cohort of 75 nonsyndromic individuals with LVNC detected three mutations, including one truncation mutant, one frameshift null mutation, and a single missense mutant. In addition, in a series of cardiac biopsies from 131 individuals with DCM, we found 5 individuals with 4 previously unreported nonsynonymous variants in the coding region of PRDM16. None of the PRDM16 mutations identified were observed in more than 6,400 controls. PRDM16 has not previously been associated with cardiac disease but is localized in the nuclei of cardiomyocytes throughout murine and human development and in the adult heart. Modeling of PRDM16 haploinsufficiency and a human truncation mutant in zebrafish resulted in both contractile dysfunction and partial uncoupling of cardiomyocytes and also revealed evidence of impaired cardiomyocyte proliferative capacity. In conclusion, mutation of PRDM16 causes the cardiomyopathy in 1p36 deletion syndrome as well as a proportion of nonsyndromic LVNC and DCM.
    The American Journal of Human Genetics 06/2013; · 11.20 Impact Factor
  • Heart (British Cardiac Society) 05/2013; 99(suppl 2):A136-A137. · 5.01 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Obesity is a major risk factor for cardiometabolic disease, but the effect of body composition on vascular aging and arterial stiffness remains uncertain. We investigated relationships among body composition, blood pressure, age, and aortic pulse wave velocity in healthy individuals. Pulse wave velocity in the thoracic aorta, an indicator of central arterial stiffness, was measured in 221 volunteers (range, 18-72 years; mean, 40.3±13 years) who had no history of cardiovascular disease using cardiovascular MRI. In univariate analyses, age (r=0.78; P<0.001) and blood pressure (r=0.41; P<0.001) showed a strong positive association with pulse wave velocity. In multivariate analysis, after adjustment for age, sex, and mean arterial blood pressure, elevated body fat% was associated with reduced aortic stiffness until the age of 50 years, thereafter adiposity had an increasingly positive association with aortic stiffness (β=0.16; P<0.001). Body fat% was positively associated with cardiac output when age, sex, height, and absolute lean mass were adjusted for (β=0.23; P=0.002). These findings suggest that the cardiovascular system of young adults may be capable of adapting to the state of obesity and that an adverse association between body fat and aortic stiffness is only apparent in later life.
    Hypertension 04/2013; · 6.87 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: AIMS: Echocardiographic studies have shown that left atrial volume (LAV) predicts adverse outcome in small heart failure (HF) cohorts of mixed aetiology. However, the prognostic value of LAV in non-ischaemic dilated cardiomyopathy (DCM) is unknown. Cardiovascular magnetic resonance (CMR) allows accurate and reproducible measurement of LAV. We sought to determine the long-term prognostic significance of LAV assessed by CMR in DCM. METHODS AND RESULTS: We measured LAV indexed to body surface area (LAVi) in 483 consecutive DCM patients referred for CMR. Patients were prospectively followed up for a primary endpoint of all-cause mortality or cardiac transplantation. During a median follow-up of 5.3 years, 75 patients died and 9 underwent cardiac transplantation. After adjustment for established risk factors, LAVi was an independent predictor of the primary endpoint [hazard ratio (HR) per 10 mL/m2 1.08; 95% confidence interval (CI) 1.01-1.15; P = 0.022]. LAVi was also independently associated with the secondary composite endpoints of cardiovascular mortality or cardiac transplantation (HR per 10 mL/m2 1.11; 95% CI 1.04-1.19; P = 0.003), and HF death, HF hospitalization, or cardiac transplantation (HR per 10 mL/m2 1.11; 95% CI 1.04-1.18; P = 0.001). The optimal LAVi cut-off value for predicting the primary endpoint was 72 mL/m2. Patients with LAVi >72 mL/m2 had a three-fold elevated risk of death or transplantation (HR 3.00; 95% CI 1.92-4.70; P < 0.001). LAVi provided incremental prognostic value for the prediction of transplant-free survival (net reclassification improvement 0.17; 95% CI 0.05-0.29; P = 0.002). CONCLUSIONS: LAVi is a powerful independent predictor of transplant-free survival and HF outcomes in DCM. Assessment of LAV improves risk stratification in DCM and should be incorporated into routine CMR examination.
    European Journal of Heart Failure 03/2013; · 5.25 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Risk stratification of patients with nonischemic dilated cardiomyopathy is primarily based on left ventricular ejection fraction (LVEF). Superior prognostic factors may improve patient selection for implantable cardioverter-defibrillators (ICDs) and other management decisions. To determine whether myocardial fibrosis (detected by late gadolinium enhancement cardiovascular magnetic resonance [LGE-CMR] imaging) is an independent and incremental predictor of mortality and sudden cardiac death (SCD) in dilated cardiomyopathy. Prospective, longitudinal study of 472 patients with dilated cardiomyopathy referred to a UK center for CMR imaging between November 2000 and December 2008 after presence and extent of midwall replacement fibrosis were determined. Patients were followed up through December 2011. Primary end point was all-cause mortality. Secondary end points included cardiovascular mortality or cardiac transplantation; an arrhythmic composite of SCD or aborted SCD (appropriate ICD shock, nonfatal ventricular fibrillation, or sustained ventricular tachycardia); and a composite of HF death, HF hospitalization, or cardiac transplantation. Among the 142 patients with midwall fibrosis, there were 38 deaths (26.8%) vs 35 deaths (10.6%) among the 330 patients without fibrosis (hazard ratio [HR], 2.96 [95% CI, 1.87-4.69]; absolute risk difference, 16.2% [95% CI, 8.2%-24.2%]; P < .001) during a median follow-up of 5.3 years (2557 patient-years of follow-up). The arrhythmic composite was reached by 42 patients with fibrosis (29.6%) and 23 patients without fibrosis (7.0%) (HR, 5.24 [95% CI, 3.15-8.72]; absolute risk difference, 22.6% [95% CI, 14.6%-30.6%]; P < .001). After adjustment for LVEF and other conventional prognostic factors, both the presence of fibrosis (HR, 2.43 [95% CI, 1.50-3.92]; P < .001) and the extent (HR, 1.11 [95% CI, 1.06-1.16]; P < .001) were independently and incrementally associated with all-cause mortality. Fibrosis was also independently associated with cardiovascular mortality or cardiac transplantation (by fibrosis presence: HR, 3.22 [95% CI, 1.95-5.31], P < .001; and by fibrosis extent: HR, 1.15 [95% CI, 1.10-1.20], P < .001), SCD or aborted SCD (by fibrosis presence: HR, 4.61 [95% CI, 2.75-7.74], P < .001; and by fibrosis extent: HR, 1.10 [95% CI, 1.05-1.16], P < .001), and the HF composite (by fibrosis presence: HR, 1.62 [95% CI, 1.00-2.61], P = .049; and by fibrosis extent: HR, 1.08 [95% CI, 1.04-1.13], P < .001). Addition of fibrosis to LVEF significantly improved risk reclassification for all-cause mortality and the SCD composite (net reclassification improvement: 0.26 [95% CI, 0.11-0.41]; P = .001 and 0.29 [95% CI, 0.11-0.48]; P = .002, respectively). Assessment of midwall fibrosis with LGE-CMR imaging provided independent prognostic information beyond LVEF in patients with nonischemic dilated cardiomyopathy. The role of LGE-CMR in the risk stratification of dilated cardiomyopathy requires further investigation.
    JAMA The Journal of the American Medical Association 03/2013; 309(9):896-908. · 29.98 Impact Factor
  • Source
  • [Show abstract] [Hide abstract]
    ABSTRACT: The mechanisms underlying coronary microvascular remodeling and dysfunction, which are critical determinants of abnormal myocardial blood flow regulation in human hypertension, are poorly understood. The spontaneously hypertensive rat (SHR) exhibits many features of human hypertensive cardiomyopathy. We demonstrate that remodeling of intramural coronary arterioles is apparent in the SHR already at 4 weeks of age, i.e. before the onset of systemic hypertension. To uncover possible genetic determinants of coronary microvascular remodeling, we carried out detailed histological and histomorphometric analysis of the heart and coronary vasculature in 30 weeks old SHR, age-matched Brown Norway (BN-Lx) parentals and BXH/HXB recombinant inbred (RI) strains. Using previously mapped expression quantitative trait loci (eQTLs), we carried out a genome-wide association analysis between genetic determinants of cardiac gene expression and histomorphometric traits. This identified 36 robustly mapped eQTLs in the heart which were associated with medial area of intramural coronary arterioles [false discovery rate (FDR) ~5 %]. Transcripts, which were both under cis-acting genetic regulation and significantly correlated with medial area (FDR <5 %), but not with blood pressure indices, were prioritized and four candidate genes were identified (Rtel1, Pla2g5, Dnaja4 and Rcn2) according to their expression levels and biological functions. Our results demonstrate that genetic factors play a role in the development of coronary microvascular remodeling and suggest blood pressure independent candidate genes for further functional experiments.
    Archiv für Kreislaufforschung 01/2013; 108(1):316. · 7.35 Impact Factor
  • Source
    Journal of Cardiovascular Magnetic Resonance 01/2013; 15(Suppl 1):E2. · 4.44 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Next-generation sequencing (NGS) provides an unprecedented opportunity to assess genetic variation underlying human disease. Here, we compared two NGS approaches for diagnostic sequencing in inherited arrhythmia syndromes. We compared PCR-based target enrichment and long-read sequencing (PCR-LR) with in-solution hybridization-based enrichment and short-read sequencing (Hyb-SR). The PCR-LR assay comprehensively assessed five long-QT genes routinely sequenced in diagnostic laboratories and “hot spots” in RYR2. The Hyb-SR assay targeted 49 genes, including those in the PCR-LR assay. The sensitivity for detection of control variants did not differ between approaches. In both assays, the major limitation was upstream target capture, particular in regions of extreme GC content. These initial experiences with NGS cardiovascular diagnostics achieved up to 89 % sensitivity at a fraction of current costs. In the next iteration of these assays we anticipate sensitivity above 97 % for all LQT genes. NGS assays will soon replace conventional sequencing for LQT diagnostics and molecular pathology.
    Journal of Cardiovascular Translational Research 01/2013; 6(1). · 3.06 Impact Factor

Publication Stats

2k Citations
741.85 Total Impact Points


  • 1999–2014
    • Imperial College London
      • • Division of Cell and Molecular Biology
      • • Faculty of Medicine
      Londinium, England, United Kingdom
    • National Heart, Lung, and Blood Institute
      Maryland, United States
  • 2009–2013
    • MRC Clinical Sciences Centre
      London Borough of Harrow, England, United Kingdom
  • 2012
    • Royal Brompton and Harefield NHS Foundation Trust
      Harefield, England, United Kingdom
  • 2008–2010
    • Max-Delbrück-Centrum für Molekulare Medizin
      Berlín, Berlin, Germany
  • 2006–2009
    • Beth Israel Deaconess Medical Center
      • CardioVascular Institute
      Boston, MA, United States
  • 2002–2005
    • Massachusetts General Hospital
      • Cardiovascular Research Center
      Boston, MA, United States
  • 2002–2004
    • Harvard Medical School
      • Department of Medicine
      Boston, Massachusetts, United States