Inferring the gene coevolution network from deep comparative genomics

It is not surprising that a missing or damaged gene can be detrimental to the organism possessing it. It is more surprising that having extra copies of genes can also be harmful, as is the case with Down syndrome, where an extra, undamaged, copy of chromosome 21 is responsible for the syndrome’s symptoms. Geneticists believe that these extra genes are harmful because they encode proteins that need to work in concert with the proteins made by genes on other chromosomes. If these interacting proteins are not present in the correct numbers relative to each other, they can disrupt cellular activities. Unfortunately, no simple experiments can identify all the pairs of genes that need to work in balance with each other. Curiously however, there are also cells that have entire extra copies of the genome in their cells, possessing not two but four or more copies. Despite the extra copies, these polyploid cells are healthy and can be found in many forms of life. In fact, over evolutionary time, several important groups of organisms, including the vertebrates and all flowering plants, have undergone such polyploidy events and transmitted the resulting extra genes to their descendants. Importantly though, not all of the extra (or duplicated) genes are kept. As a result, the patterns of gene loss and survival seen after a polyploidy can be used to identify the pairs of genes that have dosage interactions with each other. In other words, by studying evolution after polyploidy across many different genomes, it will be possible to identify the interactions driving effects like those of Down syndrome, and to shed light on gene interactions in general. This project will use evolutionary modeling software developed by the investigators to analyze duplicate gene loss and retention after more than a dozen independent polyploidies. The team will improve this tool by adding the capacity to model polyploidies that occurred “on top of” other polyploidies. Then, using more than 60 individual genomes, the investigators will infer a network of genes where the connections between those genes represent dosage interactions. In this framework, pairs of interacting genes will be those for which differences in copy number between the pair is evolutionarily costly. With the network in hand, the researchers will use existing large-scale datasets on physical interactions between proteins, regulatory connections between genes and other biochemical information to understand why certain pairs of genes need to be maintained in similar copy number in order to function correctly. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.