Dovetail Genomics’ Vampire Bat Genome Sheds Light on the Evolution of Blood Feeding


Dovetail Genomics, utilizing their proprietary ChicagoTM sequencing libraries and HiRise™ scaffolding pipeline, have produced a reference-quality assembly of the genome of the common vampire bat. The high-quality and highly-contiguous (27 Mbp scaffold N50) assembly will underpin studies of vampire bats’ extraordinary evolutionary adaptations.


Tom Gilbert and his group at the University of Copenhagen study vampire bats and the evolution of blood-feeding diets. Such diets present substantial biological challenges due to scarcity of critical nutrients in blood. Drinking blood requires highly specific adaptations, and the Gilbert group aims to shed light on the genomic basis of those adaptations. These include not only the physiological ability to obtain the needed nutrition from blood, but also those required to locate and obtain blood from the bats’ prey. To understand these phenomena, the Gilbert group has chosen as their subject Desmodus rotundus, the common vampire bat.


To tackle the analyses needed to understand vampire bats’ adaptations, the Gilbert group required a genome assembly. Many relevant analyses, including synteny-aided ortholog determination, scans for positive selection, and comparative genomics require a high-quality reference genome. Prior to the Gilbert group’s collaboration with Dovetail, no such assembly existed. Another motivation for producing a truly reference-quality assembly is the need for such a resource within the larger bat research community. Because of the likelihood that the genome produced for the GIlbert group’s studies will be adopted by the community, it was important that the result be of the highest possible quality.


Prior to engaging with Dovetail, the Gilbert group endeavored to produce their own draft assembly. The data for that assembly were generated by the BGI and included shotgun and mate-pair libraries with a mix of insert sizes up to 10 kbp. These data were assembled with SOAPdenovo to yield a 2 Gbp genome with an impressive scaffold N50 of 5.5 Mbp. However, because of the importance of the genome to the community, and to their own efforts, the Gilbert group sought to further improve the genome, which led them to Dovetail.


To begin, the Gilbert group provided vampire bat DNA to Dovetail by way of the San Diego Frozen Zoo. Dovetail then produced several Chicago libraries, which were assessed for quality, and the best of them chosen for deep sequencing. The selected libraries were sent for sequencing with 2×100 paired-end reads. Sequencing yielded 491 million Chicago read pairs (~2 HiSeq 2500 lanes), corresponding to approximately 220X physical coverage of the genome.


Next, Dovetail began scaffolding using the Gilbert group’s initial 5.5 Mbp N50 assembly as input. Because the Chicago sequence data was collected in stages, Dovetail performed two separate HiRise scaffolding runs, the first of which contained approximately half the final amount of data (242M Chicago read pairs), corresponding to ~83X coverage. This first round of scaffolding produced a two-fold improvement in N50, from 5.5 Mbp to 10 Mbp. The second and final round of scaffolding utilized the full Chicago dataset to achieve a five-fold improvement over the original assembly, from 5.5 Mbp to 27 Mbp. Remarkably, the N90 was even further improved, increasing ten-fold from ~1 Mbp in the original assembly to ~10 Mbp in the final assembly.




The final genome assembly is now in the hands of the Gilbert group. They are currently engaged in validation of the assembly, but have enough confidence in the result that they have already begun new analyses of copy number variation and ortholog discovery. This valuable resource will soon be available to the larger bat and genomics research communities, where it will be foundational in revealing the genomic basis of the many intriguing adaptations embodied by vampire bats.