We asked two simple (but still unresolved) questions. Does adaptation to one environment lead to repeatable fitness changes in other conditions? And if so, can we predict these changes based on the similarity between environments?
-
-
Prikaži ovu nit
-
We evolved 20 yeast populations in 11 lab conditions and measured their fitness across multiple other environments. At first, the data looked like a mess: different populations evolved in the same condition sometimes had very different pleiotropic profiles.
Prikaži ovu nit -
Eventually, we realized this was a feature, not a bug. Different populations acquired different mutations which had different fitness effects in other conditions. Evolution is stochastic after all.
Prikaži ovu nit -
One interesting example of randomness: some populations suffered massive pleiotropic fitness losses but others did not. These losses were caused by mutations in the yeast killer virus (see a new related preprint by
@gregoryianlang https://www.biorxiv.org/content/10.1101/700302v1 …)Prikaži ovu nit -
Although pleiotropic outcomes of evolution were unpredictable in any one population, when we looked at the ensemble of populations evolved in the same condition, we found three clear patterns.
Prikaži ovu nit -
Pattern 1: populations ecologically specialize, i.e., individuals evolved in an environment are usually more fit in that environment than those evolved elsewhere. (There are some interesting exceptions, possibly explained by Mikhail Tikhonov’s new theory https://www.biorxiv.org/content/10.1101/711531v1 …).
Prikaži ovu nit -
Pattern 2: On average, individuals pleiotropically gain more fitness in conditions that are similar to the home environment and gain less fitness or lose fitness in conditions that are less similar to the home environment.
Prikaži ovu nit -
This is very intuitive, but, to my knowledge, has never been directly shown before. If I am wrong, please respond with references. If this is true more generally, we can predict average pleiotropic outcomes of evolution based on environmental similarity.
Prikaži ovu nit -
Pattern 3: Some populations gained fitness across many conditions (generalists) and some lost fitness in many conditions (specialists), with a continuum in between.
Prikaži ovu nit -
We found no “real” physiological trade-offs, i.e., our populations are not close to the Pareto front. Nevertheless, specialization happens likely because mutations with pleiotropic costs are more abundant than “universally” beneficial mutations.
Prikaži ovu nit -
Btw,
@erjerison, is an amazing scientist with a very rare combination of skills (theory, experiment, instrument building); now postdoc in the Quake lab working on something totally different. And on the job market!Prikaži ovu nit
Kraj razgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.