Integrative study of convergent genomic evolution across three animal groups – CONVERGENOMIX

The eyes of octopuses and vertebrates, the wings of birds, bats and insects, marsupial mammals of Australia and placental mammals of the old world are famous examples of convergent evolution in the history of animals. These examples are characteristic of a phenomenon that is widespread, yet very poorly understood. In fact, in general, we don’t know the genomic origins of phenotypic convergence. For instance, we don’t know how much of the evolution of convergent phenotypes is due to the appearance of genomic changes that are convergent themselves. Answering this question would allow measuring to what extent the evolution of new characters is constrained by a given genomic context, and therefore whether evolution can be predicted. We will identify the types of genomic changes that underlie convergent phenotypic evolution by studying three distant animal groups, each showing repeated convergences.
The search for the genomic causes of phenotypic convergence can now be undertaken thanks to technical progresses in DNA/RNA sequencing. Yet, the first studies that have been published raise more questions than they provide answers. In particular, some studies find massive amounts of convergent genomic changes correlated with convergent phenotypic changes, while at the same time others find the opposite: no more genomic convergences in the presence of phenotypic convergence than when the phenotypes are different. This paradoxical situation has two sources. First, ill-fitted data sets, where very few events of convergence are studied, when many more should be included to distinguish the anecdotal from the systemic and link genomic and phenotypic convergences. Second, unproven ad hoc methods, applied only once on a single data set.
We will remedy those two problems by generating data sets in three groups in which phenotypic convergences are numerous and by developing an ensemble of robust and powerful computational methods. We will generate similar transcriptomic data sets for each of our three groups of animals to study convergent evolution in sequence substitutions, insertions, deletions, gene duplications and losses, gene expression levels and timelines. Then we will apply the same bioinformatic pipelines to each of them. This way, we will identify among the ensemble of changes those that are linked to the convergent evolution of phenotypes. These studies in three distinct groups of animals will be used to answer a few outstanding questions, including: how many ways to evolve a phenotype are there? What is the impact of non-adaptative processes on convergent genomic evolution? How does convergence in gene sequence relate to convergence in gene expression? How predictable is convergent genomic evolution?
Our consortium gathers 4 research groups with complementary sets of expertise. 3 teams are each established specialists of a group of animals showing a remarkable set of phenotypic convergences: convergences of insects from life on freshwater surface to sea water surface, convergences of surface crustaceans to underground caves, convergences of rodents living in mesic environments to arid environments. The fourth group is expert in evolutionary genomics and bioinformatics.
This project is fully realistic because partial data sets for each of the groups are already available, and were used to validate some of our main methodological ideas that will be implemented in our bioinformatic analysis pipelines. The use of three distinct data sets will allow generalizing the conclusions obtained based on the comparative study of all of them to draw hypotheses applicable to convergent phenotypic evolution at the scale of all animals. The data generated and the methods validated on all three data sets will be made available under a public license to further our understanding of the genomic sources of phenotypic evolution.