Approved Research

Mapping novel cellular phenotypes to clinical outcomes via genetic determinants.

Lay summary

Although the past nearly 15 years of genomewide association studies have resulted in a large list of portions of the genome associated with disease, there has been no clear way to map these to therapies. Very often, the problem is that the association itself starts with a complex poorly defined disease (such as heart failure) and maps it to a complex, locus that includes many genes.  Experimental science, which is needed to bridge such associations to actual therapies, requires strong hypotheses regarding genes, tissues, timing, and some quantitative measurable phenotype. Such strong hypotheses have not emerged from this type of work.

Our group has taken a different approach, building new measurements of normal and abnormal biology using whole blood isolated from many thousands of actual patients and relating them to human genetic variation as well as to disease. In this case we have very strong hypotheses: we know what the cells are, we know what the environmental stresses are, and we know which clinical populations show the largest effect.  Our new measurements of cell behavior are very strongly associated with common variability in genetic makeup from person to person. Finally, because we fully control the assays, we can now study them in much greater detail using all the powerful tools of experimental biology, in the very same system that was used for genetic discovery. 

The final step in this work is to show that those same genetic variants that determine these new assays also influence disease. Fortunately, this is different from the typical way such associations are performed - as we have very specific, strong hypotheses about which patients to study, which genetic variants to study, and the mathematical models needed to study them. We anticipate that this approach, combining our own experimental work with an invaluable resource like UKBB will result in novel therapies, accompanying diagnostics, and clear patient populations that are likely to benefit.