Centurion: joint identification of centromere locations in yeasts using Hi-C

Although centromeres are essential for life and are the subject of extensive research, centromere locations in yeast genomes are difficult to infer, and in most species they are still unknown. Recently, the chromatin conformation assay Hi-C has been re-purposed for diverse applications, including de novo genome assembly, deconvolution of metagenomic samples, and inference of centromere locations. We describe a method, Centurion, that jointly infers the locations of all centromeres in a single yeast genome by exploiting the centromeres’ tendency to cluster in 3D space. We first demonstrate the accuracy of Centurion in identifying known centromere locations from high coverage Hi-C data of budding yeast and a human malaria parasite. We then use two metagenomic samples with relatively low coverage Hi-C data to infer centromere locations for each chromosome in 14 different yeast species. For yeasts with large centromeres (e.g., S. pombe) Centurion predicts the exact centromere locations. For seven yeasts with point centromeres, Centurion predicts most of the centromeres at an average of 5~kb distance from their known locations. Finally, we predict centromere coordinates for six yeast species that currently lack centromere annotations. These results suggest that Centurion can be used for centromere identification for a large number of yeast species, even with a limited amount of Hi-C sequencing.


Download Centurion 0.1 here or fork the code on github.


If you have any questions or suggestions, please email nelle dot varoquaux at ensmp dot fr, or open a ticket on Github