Genome Duplication Reshapes Salmon Evolution

[SOFIA] Okay, this is the good stuff! We're kicking off today with a fascinating piece of evolutionary biology that hit the press this week, all about how something as massive as duplicating an entire genome can completely reshape a family of organisms – in this case, the salmon.

[DANIEL] Hm — worth pausing on the scale here, because Sofia's right that it's massive. Most of the changes evolution works with are single-nucleotide, one letter on one page, but a whole genome duplication copies the entire book at once — every gene, suddenly present in two copies. And in salmon that's not hypothetical: the whole family descends from an ancestor that went through exactly that, so the interesting question is what happens to all those redundant genes afterward.

[SOFIA] Exactly! And it wasn't just *a* duplication; the salmon ancestor experienced what's called a 'tetraploidization event' about 80 to 100 million years ago, meaning its entire genome doubled, going from two sets of chromosomes to four.

[DANIEL] So then you've got four copies of everything, and the cell has to sort that out — most of those duplicates get silenced or lost over time, but a fraction get kept and repurposed, which is where the disease-resistance and reproduction traits come in. What I'd want to see from the actual work is how they distinguished a gene that picked up a genuinely new function from one that's just a passenger still drifting toward pseudogene status — that's the hard call in this kind of analysis.

[SOFIA] Well, that's what's so cool about this new analysis — they looked at the *rate* of evolution in those duplicated genes, essentially asking which ones stuck around and diversified fastest, and found that genes involved in things like immunity and reproductive development were evolving at a much quicker clip post-duplication.

[DANIEL] Faster evolution in the immune and reproductive genes is exactly the signature you'd hope for — but "quicker clip" can mean two very different things, and I'd want to know which they measured: relaxed constraint, where a redundant copy just drifts because nothing's pruning it, versus positive selection actually pushing a new function. If they've got the dN/dS ratios showing selection rather than just neutral drift, then this holds up as adaptation and not bookkeeping.

[SOFIA] Okay, this is the good stuff! That's exactly what they got into – they looked at a process called "subfunctionalization," where the duplicated genes take on specialized roles, and saw clear evidence of positive selection driving those immune and reproductive genes to diversify, rather than just drifting or getting silenced.

[DANIEL] Hm — one thing I'd flag: subfunctionalization and positive selection aren't quite the same story. Subfunctionalization is the two copies splitting the ancestral job between them, which is often neutral; positive selection driving a genuinely new function is neofunctionalization. If they're seeing both in the immune and reproductive genes, that's a strong result — but I'd want the dN/dS numbers and the copy counts before I'd call it settled.

[SOFIA] That's a really important distinction, Daniel, and this study actually *does* get into that, showing how those duplicated genes in salmon often go on to specialize, with some copies evolving completely new functions through positive selection, especially in those defense and reproduction pathways.

[DANIEL] And that's the payoff for anyone who works on non-model organisms — salmon are basically a natural experiment in what a genome does with a spare copy of every gene, and the redundancy is what gave selection the raw material to build new immune and reproductive functions. I'd still want to read the dN/dS tables myself before I'm fully sold, but if they hold, this is a clean look at duplication as an engine of adaptation — Sofia, where are we headed next?