Crazy Sheep DNA from Medieval Manuscripts Project: Update

For quite a number of years now I have been involved with a long-term interdisciplinary project to extract DNA from medieval parchment and use the information so retrieved to help figure out relationships between manuscripts.

The project started back in 2005 or so, and for a while was more speculation and planning than anything else. Then I got some summer funding and trained a student in paleography while she simultaneously learned how to do Polymerase Chain Reaction work in Biology (she was a Bio major who has since gone on to medical school). A second student followed up and extracted more DNA from medieval parchment (I bought a leaf and a fragment at Kalamazoo for this purpose).

But we always had fears about contamination: although we did get what looked to be ovine DNA, we had some issues with our controls, making it seem like we might have had a contaminated room or contaminated reagents. Which is why I didn't rush into print or publicity like some other groups (who shall remain nameless, and really could have been more polite about priority and public discussion, but let that pass): we had the data a long time ago, but we just couldn't be sure.

Today, however, we confirmed as closely as we could that the apparent contamination was from human, not ovine, DNA, and so the previous results (success in extraction) will most likely stand up. We are running tests again, and today got confirmation that we got sheep DNA from modern parchment (thanks, Pergamana!). The two next steps are to go back to the medieval parchment and to see how small a sample we can use. Our results on modern parchment were done on 25mg of material: about 3x5 mm, which is too large to use on a lot of manuscripts (though might be useful for binding fragments, etc.). Librarians just aren't going to let you snip off even a corner that small (and rightly so). However, if we can reduce the necessary sample size to 2.5 mg, that's another story.

At the same time, we need a good way of getting samples that isn't destructive. That has been the project of a bunch of amazing engineering students at Northwestern University, with whom I've been working to design a Sheep DNA Extractor (a more accurate description is that I throw out crazy ideas and give them sketches on cocktail napkins, and they make cool prototypes that seem to work).

Hopefully these two strands of effort will come together in the next two weeks, when we finally get some data on the minimum sample size and the Northwestern students then choose which of their cool designs work.

The next step will be designing the database into which this information will go and a user interface both for inputting the information and for retrieving and manipulating it. That's where the Computer Science majors at Wheaton come in: for their senior seminar project, I am their "client" and they are going to work with me to design and program all the software.

The overall plan is that at the end of this year, we can present an integrated project, from procedures for sample-taking that is not visibly destructive, to the biological work to extract the DNA, to the database for storing it, to the software which looks for patterns and then allows readers to access the information.

This is one of the harder things that I have done, because I'm coordinating a number of disparate efforts, each of which has a different timetable, different problems and and different priorities: it reminds me of trying to cook a big meal, rushing from pot to pan to over to chopping block, juggling things frantically. But that's also what makes it fun.

And who knows, once I write this all up and present the still-secret extraction method and the interface and the plan for going forward, maybe some funding body or wealthy philanthropist will give the project some support.

In any event, and even before finishing (and of course any part or parts could still crash and burn), this has definitely been worth it, because it has given me an excuse to hang out in the Science Center and worth with biologists, computer scientists and engineers. All because I gave Scott McLemee a crazy answer to a question in 2005.


(by the way, as far as I know the first person to have the idea of tracing manuscripts through sheep DNA was Greg Rose, who told me about his idea in September 2001).