Jason Ernst LECIF

Learning a genome-wide score of human–mouse conservation at the functional genomics level | Nature Communications

Soo Bin Kwon & Jason Ernst 

Mouse is a widely used model organism in biomedical research for studying a wide range of questions with direct relevance to understanding human disease. However, the extent to which mouse is an effective model can vary drastically depending on the specific question being studied. This motivates the need to better identify circumstances where mouse would more likely be an effective model organism for human. In particular, about 40% of bases in the human genome can be mapped to a corresponding location in the mouse genome based on DNA sequence similarity. However, it is often unclear if these corresponding locations actually have similar roles that for example might be associated with the same disease in both organisms.

To identify pairs of corresponding locations in human and mouse with additional evidence of having similar roles, Kwon and Ernst developed a new computational method, LECIF (Learning Evidence of Conservation from Integrated Functional genomic annotations). LECIF is a machine learning method applied to data derived from thousands of experiments that map various biochemical activities along the genomes in diverse human and mouse tissues. LECIF effectively learns a score for evidence of conservation between human and mouse at the functional genomics level, providing additional evidence of conservation beyond sequence. The resulting score thus highlights locations in the human and mouse genomes for which findings from mouse model research are more likely to transfer to human biology. 

"This work was funded in part by the US National Institutes of Health, US National Science Foundation, Kure It Cancer Research, and the Rose Hills Foundation."

Media Contact: 

Leticia Ortiz | Marketing & Communications | Building a community around data science in biomedicine