Quick thread on evolution, mostly so I can refer to it from elsewhere.

Adaptive evolution by natural selection requires two fundamental materials: diversity and selection pressure

First, let's define some terms:

We'll be thinking about *replicators*: things that can make (possibly imperfect) copies of themselves

Diversity, in this context, means that not all of the replicators are identical. There are different *types*

Selection pressure means that for whatever reason, some of those types produce new copies of themselves more readily than others

Evolution, in this context, means that over time some types of replicators become more common, and other types become less common. e.g. right now the B.1.1.7 variant of SARS-CoV-2 is becoming more common in the UK than the ancestral variant. That's evolution.

But why did I say *adaptive evolution* at the start?

One reason that one type can become more relatively common than another over time is that it is favored by selection pressure—something about it or about the environment means that it *tends to replicate better*

That kind of evolution—becoming more common because you're favored by natural selection—is called "adaptive evolution"

But while it's the kind of evolution that we focus on in a lot of popular discussion, it's not the only possible kind

Sometimes, evolution happens when two types are are more or less equal at replicating, but one gets lucky and the other gets unlucky. This is called neutral or nearly neutral evolution.

You may recall that when B.1.1.7 was first identified, some people wondered if it really was more transmissible, or if it was just lucky.

More transmissible: adaptive evolution
Just lucky: neutral evolution

When we ask about antigenic changes to a virus that allow it to evade our existing immunity, we're first and foremost worried about adaptive evolution. Our immunity imposes selection pressure on the virus, because viruses that better evade that immune will tend to replicate more

But NB that for the virus to escape our immunity, that selection pressure is not enough.

We also need diversity. A new virus type that better evades immunity needs to exist or be created

And here's where we get into the weeds. Because in reality, you can't always completely decouple the question of diversity from the question of selection pressure

To stick with viruses, within a single infected individual, killing off viruses with an immune response creates selection pressure for immune escape, but it also limits the virus's ability to generate diversity through mutation

And it gets more complicated because some kinds of immunity may affect all the viruses equally (reducing diversity generation without creating meaningful selection pressure), while others affect some more than others (creating selection pressure)

@colinrussell, @veli_pet, colleagues, and I have modeled what we think is a pretty cool potential example of this in flu antigenic evolution: https://twitter.com/dylanhmorris/status/1326906383569186816

So when you think about adaptive evolution, remember that both diversity and selection are necessary, but also remember that certain forces can affect both the level of diversity and the level of selection pressure

/end