Share This article

Sequencing DNA is very, very simple: there’s a molecule, you look at it, you write down what you find. You’d think it would be easy — and it is. The problem isn’t looking in and checking the chemical identity of each link in the chain of a molecule of DNA, it’s checking those identities tens of millions of times while making essentially no mistakes. That is what’s hard, but the nature of DNA is such that if you’ve only got 95% of the correct sequence, you might as well have nothing at all. So how do scientists actually read the blueprints of biology, and with them build a huge proportion of modern medicine and biotechnology?

It all started, more or less, with a guy named Frederick Sanger. Sanger created an ingenious method of reading a DNA molecule, which involved using a specialized version of DNA bases called dDNA, or di-deoxy-ribonucleic acid. The ‘di’ refers to the fact that dDNA bases are without both of the -OH groups found on RNA bases, while normal deoxy-ribonucleic acid (DNA) still have one. In normal DNA bases, this single -OH group acts as the attachment point for the next link in the chain of a DNA molecule. Without one of its own, dDNA bases can’t form DNA’s characteristic chains, so they end any chain-growth process when they’re incorporated into a growing DNA strand. Sanger realized he could exploit this tendency of dDNA bases to stall any chain-elongation process to see the sequence of the chain itself.

The speed of DNA sequencing has been increasing exponentially, but can that trend continue?

Let’s do a quick thought experiment: Let’s say I have a 4-base DNA molecule with the sequence ATGC, though I don’t know that sequence and I’d like to. I know that DNA can be made to replicate itself fairly easily; just heat it to the point that the double helix “melts” into two separate strands in the presence of enzymes that snap free-floating DNA bases onto them, and you’ll eventually end up with two separate double helices where you originally had one. But what if the free-floating bases being snapped onto these single strands are a mix of regular DNA bases and “terminal” dDNA bases?

Well, in that case we’d get a mixture of products, depending on where in the growing chains our fluorescently-labelled terminal dDNA bases ended up being inserted. For our ATGC molecule, some of the replicated strands would be full length and unlabelled — no dDNA base happened to get inserted at all. But we’d also end up with some one-base strands ending in the dDNA base C — just a single A-C base pair. More helpfully, we’d also get a mixture of two-base strands ending in a labelled G, three-base strands ending in a labelled T, and four-base strands ending in a labelled A. This gives us a sequence read of CGTA, meaning the original complementary sequence was ATGC.

However, even automating this process remained far too slow to allow the sort of population scale meta-analysis modern medicine and genomics were requiring. That’s where so-called “massively parallel sequencing” came in, sometimes colloquially referred to as shotgun sequencing. This basically refers to the idea that if you break a long sequence of DNA up into smaller fragments, you can simultaneously read them all. You have to read many, many copies of your overall sample, since you have to take that fragmented data and run a puzzle-like algorithm to figure out how they went together in the first place.

Selexa sequencing, simplified.

The most popular of these shotgun methods was probably Solexa, which saw DNA broken up and adhered to a glass plate. The process uses reversibly terminal bases — bases that will stall the chain-growth process for a while, until the scientists choose to unblock them and allow the next link to be added. The strict add-read-unblock cycle lets scientists take a snapshot of many millions of fragments, reading the base at the end of each one before allowing the addition of another temporarily terminal base and taking a new snapshot.

Massively parallel sequencing changed the game for genomics researchers, but it’s the step after even these techniques that could revolutionize public health by making enormous sequencing speed much more affordable and practical. There are several competing bids to do this, but they all attempt to remove the DNA replication process altogether — so-called “direct” reading of a DNA molecule without the need for messy, demanding, time-consuming reactions of DNA with enzymes.

Next-gen DNA sequencers are more than just fast — they’re practical.

The most successful of these early technologies is nanopore sequencing. This method actually feeds a strand of DNA through a pore in a conductive material. As the bases move through this nanopore, their slightly different sizes stretch the pore a characteristic amount — and that change in mechanical stress on the pore translates to a change in electrical conductivity. By reading the changes in conductivity as a strand of DNA is fed through a nanopore, these sequencers can do away with the replication reactions of old.

That will be important, as more and more DNA technologies are invented that could help aid workers in inhospitable environments, or just millions of family doctors around the world who can’t afford to run a Solexa experiment every day or so. Improving sequencing tech will open a few new research doors, but for well-funded labs the limits on sequencing are already astronomically high. At this point, the import of newer, better sequencing tech is in the ability to democratize probably the most emergent branch of the physical sciences, right now. Sequencing breakthroughs may allow new scientific insights, but more likely they’ll allow real-world application of insights we’ve had for a while.

All those articles you’ve read about the potential of personalized medicine? These sorts of sequencing breakthroughs will need to continue, to make them a reality. But unlike the graphenes and the superconductors of the world, sequencing tech almost undeniably will get there, and not slowly. So, the question now becomes not, “How do we sequence more DNA?” but rather, “What can we do with those sequences, once we’ve put them in as many hands as possible?”

Check out our ExtremeTech Explains series for more in-depth coverage of today’s hottest tech topics.