Sequencing Thylacine Mitochondrial DNA

Our method for generating the sequence data was very similar to how we sequenced woolly mammoth. One critical difference was that with the mammoth we started with a large ball of hair, whereas with the rarer thylacine specimens we had only tiny amounts of hair to work with. Thus, we were not able to enrich the thylacine material for long pieces, but instead tried to sequence even very short fragments. As a result, while our woolly mammoth sequences averaged 120 nucleotides in length, for the Smithsonian thylacine skin the average was 87.5, and for the ethanol-stored specimen it was only 67.

In addition to the problems that stemmed from working with shorter sequences, analysis of the thylacine data was complicated by the lack of a genome sequence from a close relative to use as a reference point. With mammoth, we could use the sequence from the African savanna elephant, which is the same as the woolly mammoth sequence in over 99% of the genomic positions. With the thylacine, the closest sequenced genome was from a South American opossum, Monodelphis domestica, whose genome sequence is quite different from the thylacine's.

In this regard, sequencing the thylacine was more representative of sequencing an arbitrarily chosen animal, so we were glad to address the challenge. By solving the problem of assembling the thylacine mitochondrial sequence without having a closely related "template sequence", we have moved closer to the day when sequencing museum specimens will be commonplace.