Watching an arms race one lineage at a time

The work: High-resolution lineage tracking of phage–bacterial evolution. In preparation — I’ll link the preprint here when it’s out.

Coevolution experiments usually give you a population average. You know the phage got better at infecting and the host got better at resisting, because that’s what the titres say. What you don’t get is which lineage did it, when, or what else was happening at the same time.

So we barcode both sides. A 5k unique barcodes to host & 60K unique barcodes to phage. That turns a coevolution experiment from a pair of curves into something closer to a genealogy — you can follow individual lineages on both sides of the interaction while they are still competing, rather than reconstructing the winner afterwards.

The arms race: steps and sweeps

Who wins the early race? Phage vs. Bacteria.

In the first few days of the experiment, something surprising happened. The phage showed an incredible burst of diversity, with thousands of tiny genetic "tribes" popping up almost overnight, each finding clever ways to thrive. However, within a few days, most of them crashed and disappeared, leaving just a handful of standout winners to take over the whole population. It was fascinating!

Meanwhile, the host played a much slower, longer game: instead of one dramatic burst, they showed a steady churn, with different lineages rising and falling in a tug-of-war that lasted the entire month.

A hidden demolition derby: what clonal interference actually looks like?

On Day 2 of the experiment, it looked like nearly every phage lineage had won the lottery, tens of thousands of distinct viral lineages suddenly surged with a big fitness advantage, all at once. But this wasn't ten thousand separate success stories. It was ten thousand lineages piling into the same narrow doorway at the same time, all trying to exploit the same opportunity, and only a tiny handful could actually fit through.

Rest of the story coming soon....