Scitable (and a weird piece on DNA barcoding).

I received an email about Scitable, a new online resource by Nature Education. I notice that they have a link to my 2005 paper in Nature Reviews Genetics. Overall, I think the site looks interesting. On a more curious note, I was checking out the material about comparative genomics, and came across this, um, bizarre discussion of DNA barcoding:

[Excerpt from Genomes of Other Organisms: DNA Barcoding and Metagenomics]

Partial Gene Sequences

The method of comparative genomics can be applied not just to full genome sequences, but also to single genes and gene fragments to study their function and help establish relationships among species. Indeed, a species‘ place on an evolutionary tree is a valuable predictor of the structure and function of neighboring taxa.

The current convention of describing (defining) organisms new to science and establishing their evolutionary relationships is based on total evidence; in other words, the organisms’ genetic, morphological, and ecological characters are described and analyzed against other sets of data. Taken together, these techniques can be very informative, having thus far provided us with a detailed road map of Earth’s biota. But for systematics – the study of biological diversity and common ancestry – rapid technological advances in the field of comparative genomics are both a blessing and a curse. Consider, for example, the technique called DNA bar coding, which is based on using short fragments of mitochondrial gene CO1 to uniquely identify and document animal species (Savolainen, 2005). This technique has applications across all living organisms, but the precise genetic methodology is still being developed. In addition, the debate among scientists regarding the use and the utility of DNA bar coding has been quite vociferous. On one hand, this technique brings the promise of instant species identification to a much wider community with minimal biological training. Indeed, it is hypothetically possible to carry a hand-held device out in the field and input species sequences into a rapidly expanding database; all for a fraction of the price, knowledge, and effort associated with the conventional manual method or with human-curated taxonomic identification. So what’s the catch?

One major problem with DNA bar coding is that it operates on the assumption that species have evolved in perfect percentile distances of genetic diversion. Thus, with this technique, in order for any two organisms to be deemed the same species, they must share 88-98% of genetic code at the chosen CO1 mitochondrial gene fragment (Savolainen, 2005). The exact suggested threshold has to be characterized for each group, and neither the threshold nor the groups have been clearly defined for most taxa. Thus, DNA bar coding has been called a “quick fix” and an oversimplification of systematics. Indeed, wide variation in the CO1 gene is found not only among species, but also within them, and even between the cells of an individual organism – a phenomenon known as mitochondrial heteroplasmy (Kmiec & Woloszynska, 2006). Furthermore, there is a broad overlap of inter- and intraspecific genetic distances among closely related species (Goldstein et al., 2000).

These issues come into focus when you consider the devastating malaria epidemic that kills one to three million people worldwide every year. The pathogens that cause malaria are protozoan parasites from the genus Plasmodium that are transmitted through the bite of mosquitoes of the genus Anopheles. Both of these animal genera contain hundreds of species, although only a few are involved in transmitting malaria in humans. Recent genetic studies of the symbiotic bacteria in the midgut of the Anopheles stephensi mosquitoes have yielded promising results: Enterobacter agglomerans bacteria were genetically engineered to display two anti-Plasmodium effector molecules that kill the parasite before it is transmitted to humans (Riehle et al., 2007). Now consider the genetic and physiological differences between the wild-type and genetically modified Anopheles stephensi mosquitoes: they are still the same species by all major standards of species definition, yet what a difference it would make for humankind if the Plasmodium-resistant genetically modified strain were dominant. This example highlights the importance of studying genomes and biological associations of the narrowest niches of life. It also underlines the vital potential for the unpredictable outcomes of genome sequencing-major advances are often made using information generated for completely unrelated reasons.

Ok. The last paragraph has nothing to do with DNA barcoding. As to the critique of DNA barcodes, I find it odd that the author a) does not cite any papers by people who do DNA barcoding, and b) cites a paper from 2000 (i.e., 3 years before DNA barcoding began) as showing that sequences overlap. Nevermind that the paper a) does not discuss COI data, and b) is co-authored by Rob DeSalle and Alfried Vogler (my two postdoc advisors), neither of whom is against the use of DNA in species identifications (though both would prefer cladistic methodology).

The bias is obvious, but at least they could have included some proper references (goodness knows there are enough vocal opponents).