Friday, September 07, 2007

Don't Trust All DNA Sequences

I've been interested in sequence errors and cloning artifacts for many years. I have a thick file of papers that have uncovered numerous examples of DNA sequences that are just plain wrong. It's one of the reasons for being at least a little bit skeptical of any unusual discoveries in DNA sequences.

Now, I'm not saying that you should take this skepticism to extremes; I'm saying that you should keep your mind open to the possibility that the data might not be real. Just because it's published in the peer-reviewed literature doesn't mean it's right.

Sandra Porter has an interesting example over on Discovering Biology in a Digital World [Digital Biology Friday: What sequences do you believe?]. She describes the process she went through when she first heard about the discovery of a β-lactamase gene [Penicillin Resistance] that was supposedly from Streptococcus pneumoniae. I'm not going to spoil the punchline by revealing it here. You'll have to read about it on Sandra's blog.

Keep this lesson in mind. It's what good science is all about.

Figure Credit: The image of the common (but old) cloning vector pBr322 is from Horton et al. 4th ed. (2006). The ampR gene encodes β-lactamase.]


  1. Supposing someone told you they had extracted human DNA from a wad of chewed leaves 800 to 2400 years old, would you be suspicious?