Wild tuco-tuco. Photo by Anand Varma, former member of my lab; http://www.varmaphoto.com/

Take a tooth. Leave it in a cave for 5000 years. Retrieve it and examine the tooth: after all that time, those seasons passing and bacteria working away, what is left of the original animal? Not a lot; but not nothing.

There remains still some DNA from the original owner of the tooth, but degraded, fragmented into little pieces, and overwhelmingly outnumbered by the DNA of all the bacteria that have grown and reproduced and died in the tooth. Finding the DNA of the original animal would be like finding a needle in a haystack—if the haystack was really big and the needle was also a piece of hay, just slightly different from all the other hay.

And yet: we can do it.

My tuco-tuco work involves taking the DNA from these very old tuco teeth—a sort of DNA soup of tuco DNA, bacterial DNA, and potentially lots of other things, including human DNA if I have been sloppy and allowed my own genetic material to contaminate the sample—and isolating and magnifying just the tuco DNA, so that I can learn things about the tucos that lived long ago. This is a challenge for many reasons: for example, we can’t really “see” the DNA until we sequence it, which is the expensive step and not something you want to do until you have all the contaminating DNA cleaned away. We are trying to find a hay-needle in a giant haystack in the dark.

And yet: we can do it.

The oddest thing about this project, to me, is the scale. None of my prior research involved so many numbers of such varying magnitudes, so many figures that are simultaneously enormous and miniscule.

Our tuco DNA fragments are about 60 base pairs long: they are twinned lengths of 60 components connected to each other like beads on a string. This is very small, for DNA: these are the remnants of DNA that has been shattered by time. It is much too small to see or to pick up with the tiniest tweezers. Suspended in liquid, which is how we keep it, it is invisible in what appears to be entirely clear water. And yet we manipulate it like a child making crafts with safety scissors and a glue stick: we glue on extra bits of DNA to the ends, we attach it to magnets and then detach it again. It is fragile and precious, and yet we regularly break it in half (“denature” it) and then watch it heal itself.

We have so little of this DNA: we measure it out in microliters, millionths of a liter, each about 1/50 of a standard drop from an eyedropper. Yet in each of those microliters we have hundreds of billions of our DNA fragments.

When we do eventually sequence our DNA, we will get tens of millions of “reads”—transcripts of DNA sequences—so many that I will need to store them on an external hard drive, even though they are only text files. We will then filter them for only the top-quality reads, throwing away a million or more pieces of data. Finally all the remaining reads will be spread among the many, many different DNA fragments we have, so that even while we have millions of reads, each fragment may have been read only two or three times.

When I think hard about this, it gives me a sort of vertigo; it feels like simultaneously counting how many trees are on the Earth and how many atoms are in a single leaf.

At the end of the lab work, though, all of these huge and tiny scales will realign with each other and my familiar scale will come back into focus: that of the organism, the individual tuco-tucos who grew these teeth and wore them down eating vegetation and dug burrows and had babies and eventually died, all those thousands of years ago. Being able to look back through time at them is a genuinely incredible thing, so perhaps it is appropriate that our ways of doing that are a bit incredible too.