Just How Good was Sykes' Mitochondrial 12S rRNA Gene at Distinguishing Species?

In his recent paper (B. C. Sykes, R. A. Mullis, C. Hagenmuller, T. W. Melton, and M. Sartori. Genetic analysis of hair samples attributed to yeti, bigfoot and other anomalous primates. Proc. R. Soc. B. 2014: 281, (1789), p. 20140161.  Available free at http://rspb.royalsocietypublishedncounting.org and at right) Prof. Bryan Sykes et al. sequenced and identified 30 hair samples from around the world, all purportedly from sasquatch, yeti, almasti, orang pendek, etc.  None could not be attributed to a known extant mammal or closely related mammals, and only one was human.  The technique, known as mitochondrial 12S rRNA (T. Melton & C. Holland.  J Forensic Sci, 2007: Vol. 52, No. 6, pp. 1305-07. Available free at http://www.mitotyping.com/page/37), uses a sequence of only 104 mtDNA bp to compare to a database of known animals, the NCBI GenBank using BLAST(TM) search software.  How can such a small sequence distinguish among species such as three different bears(American black, polar, brown), horse, dog, cow, deer, serow, sheep, raccoon, tapir, and porcupine?  How certain is the identification?

To answer these questions we took each of Sykes' 30 sequences and searched the nucleotide database at NCBI with BLAST(TM).  We found four relatively minor ambiguities:

(1) The wolf, dog, and coyote (all genus Canis). These take nuclear DNA sequences to distinguish.

(2) The domestic sheep and the Himilayan tahr (both sheep family).

(3) Some brown bears have the same sequence as most polar bears, and Ursus thebetanus japonica, i.e. the Japanese black bear, matches the majority of brown bears. These three species are all genus Ursus.

(4) The white-tailed and mule deers (both genus Odocoileus).  These deer species are known to hybridize.

Cases (1), (3) and (4) are genus specific, Canus, Ursus, and Odocoileus respectively.  Case (2) is only family specific (sheep). All other animals mentioned above gave a 100% ID match of all 104 bases to only one species, with related species giving consistently poorer matches. The lone human sample did not match any other primates nearly as well.


The case of bears is an interesting one.  Recent work (C. Lindqvist et al.  Proc. Natl. Acad. Sci. 2010: 107, (11), pp. 5053–5057.  Available free at http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2841953/) shows that they underwent relatively recent (Pleistocene) divergence, and may, in fact, still be responding to drastic climate changes, especially the brown and polar bears.  These two species are known to hybridize, a fact recognized by Sykes et al.

In their paper Sykes et al. recognize the Cases (1) and (4) ambiguities but not the Cases (2) and (3). Two samples were, however, described as intermediate between brown and polar bears. I asked Prof. Sykes about (3) but got no response. New polar and brown bear data have recently been added to GenBank, making a more detailed analysis of Case (3) possible for some Ursidae expert (not me). 
  
The conclusion (shared by Sykes et al.) is that some species have unique mitochondrial 12S rRNA sequences, and others do not.  Those that do not can be identified at the genus level (as recognized by Sykes et al.) and occasionally only the family level (e.g. the sheep above).

This would have been a useful technique for Melba Ketchum to have used on her bear (S26) and dog (S140) samples.  No way would they show human or primate mitochondrial 12S rRNA sequences.  Come on, Melba, try it.