Genetic correlation-guided mega-analysis of DO mice provides mechanistic insight and candidate genes for age-related pathologies

Diversity Outbred (DO) mice are a powerful model system for mapping
complex traits due to their high genetic diversity and mapping
resolution. However, while there are extensive tools available
for standard genetic analysis in DO mice, fewer techniques have been
implemented to facilitate integrated, cross-study analysis. Here, we
implement Haseman-Elston regression to estimate genetic correlations among
7,233 phenotypes measured across eleven independent DO mouse
studies.  We used this network of genetic correlations to cluster
phenotypes according to shared genetics, which enhanced the power to
detect quantitative trait loci (QTL). This approach empowered the
detection of 884 QTL for 383 meta-phenotypes, explaining an average of
40.36% of the total genetic variance per mega-analysis. We leveraged this
network for insights into specific areas of biology, including lifespan,
frailty, immune composition, histological and functional lung phenotypes,
and histological phenotypes of the aorta. We found the genetics of
lifespan to share limited correlation with the genetics of frailty but
stronger correlation with the genetics of immune cell composition.
Additionally, mega-analyses driven by genetic correlations identified
candidate genes (e.g., Cdkn2b) associated with degraded extracellular
matrix in the aorta. Finally, an ensemble of genetic analyses implicated
pulmonary neuroendocrine cell signaling and/or differentiation as a key
driver of multiple lung pathophenotypes.