Accounting for chimerism in demographic inference: reconstructing the history of common marmosets ( ) from high-quality, whole-genome, population-level data.

As a species of considerable biomedical importance, characterizing the evolutionary genomics of the common marmoset ( ) is of significance across multiple fields of research. However, at least two peculiarities of this species potentially preclude commonly utilized population genetic modeling and inference approaches: a high frequency of twin births and hematopoietic chimerism. We here investigate these effects within the context of demographic inference, demonstrating via simulation that neglecting these biological features results in significant mis-inference of the underlying population history. Based upon this result, we develop a novel approximate Bayesian inference approach accounting for both common twin-births as well as chimeric sampling. In addition, we present novel population genomic data from 15 individuals sequenced to high-coverage, and utilize gene-level annotations to identify neutrally evolving intergenic regions appropriate for demographic inference. Applying our developed methodology, we estimate a well-fitting population history for this species, which suggests robust ancestral and current population sizes, as well as a size reduction roughly 7,000 years ago likely associated with a shift from arboreal to savanna vegetation in north-eastern Brazil during this period.