Read on Twitter
Why do bacterial plasmids carry some genes but not others? Plasmids are less reliably inherited than the chromosome and genes are free to hop between plasmid and chromosome: so what keeps genes on plasmids? One answer with @Jana_Huisman and @sbonhoeffer: https://www.biorxiv.org/content/10.1101/2020.08.04.236455v1.abstract
The gist: modelling predicts gene location (i.e. on plasmid versus chromosome) is under positive frequency-dependent selection. This can keep moderately beneficial genes on plasmids, despite occasional plasmid loss.
PFDS leads to a priority effect: whichever form is acquired first has time to increase in frequency and thus become difficult to displace; traits that are more frequently acquired on plasmids than the chromosome will therefore be found on plasmids (not as circular as it sounds).
PFDS arises when the cost of having the gene on both plasmid and chromosome outweighs the benefit of having two copies; an example of effects of polyploidy in plasmid evolution. (See also cool study on plasmids, polyploidy and genetic dominance: https://www.pnas.org/content/117/27/15755)
I've enjoyed thinking about this work the context of a recent paper with @ChristoPhraser , @RafalMostowy et al. on whether the long-term fate of resistance genes depends on stochastic gene acquisition events.
We found that resistance frequencies in pneumococcal lineages are best understood as the deterministic outcome of selection pressures. https://advances.sciencemag.org/content/6/21/eaaz6137
But in the context of gene location, the priority effect from PFDS means that the timing of gene acquisition events sets the population onto an evolutionary path from which it then can’t deviate. In this context, eventual evolutionary outcomes are fundamentally stochastic.
