Exome and genome sequencing (GS) are increasingly conducted within large-scale human research
studies, creating questions for investigators about whether and how to return genetic findings to participants.
Yet critical knowledge gaps exist about the impact of such return on research participants and investigators,
how to efficiently identify and confirm such variants, and how to provide accurate estimates about their
penetrance, particularly among participants in population-based studies and those who are underrepresented
minorities. Empirical data on these questions are urgently needed. Recently, DNA collected from 7,603
individuals in the Framingham Heart Study (FHS, n=4,197) and the Jackson Heart Study (JHS, n=3,406) were
sequenced as part of the NHLBI Trans-Omics for Precision Medicine (TOPMed) Program, of whom 2,885 in
the FHS and 2,674 in the JHS, are living. An estimated 1-2% of these individuals carry a detrimental variant
(defined as pathogenic or likely pathogenic variants that are heterozygous for dominant conditions or bi-allelic
for recessive conditions) in one of 59 genes for actionable conditions that the American College of Medical
Genetics and Genomics recommends be returned in clinical sequencing, regardless of indication.
In Aim 1 of this study, we will develop and implement a genomic return of results process for individuals
noted to have detrimental variants in one of these 59 genes, and evaluate medical, behavioral and economic
outcomes associated with returning this information in community-based research populations. As the
interpretation of genomic results to identify true pathogenic variation is a highly labor-intensive process, in Aim
2 we will refine and apply methods for high throughput screening of FHS/JHS genomes in a manner that
retains high sensitivity for the detection of detrimental variants in an additional 4,572 disease-associated
Mendelian diseases (which will not be returned) while reducing the false discovery rate of variants that are
determined to be benign. In Aim 3 we will review available phenotype data and compare documented and self-
reported phenotypes to variant classification in participants who do, and do not, carry detrimental variants in all
4,631 Mendelian disease-associated genes. By comparing genotype and phenotype data, we will refine a new
method of estimating crude measures of aggregate penetrance and lay the groundwork for generating more
refined penetrance estimates in larger sample sizes.