How can we time a recent gene duplication event using (neutral?) DNA mutations as molecular clock?

Question

[xiaofeng.dong12]

I found some gene duplicates have a nearly 100% percent DNA sequences in a insect genome. They are not simple repeats. Are they possible to be a very very recent duplication? for examples in 10-100 years. I think the gene loss will take time. How is the duplications?

Thanks

[5heikki]

Sure, why not. You should have asked at biology though..

[Giovanni M Dall'Olio]

Is the duplication fixed in the population? In other words, did you find it in more than one individual, or just in one?

[Michael]

I took the freedom to make the title a bit more general.

[Devon Ryan]

Well a gene duplication can happen in under a day if you were able to follow single cells over time. If you're talking about population level changes, however, then 10 years would probably be too quick, even for insect.

But anyway, 5heikki is correct that you should ask this on the biology stackexchange.

[Brian Bushnell]

10 years would probably be plenty for an insect population, if it was a pesticide-resistance gene :)

[Istvan Albert]

to extend on that - one factor would be generation time (the time between consecutive generations, depending on the insect 10 years could span very many or very few generations), the type of selective pressure and well ... luck really. Sometimes you see things because they have already happened.

I recall the result of a "back on the envelope" calculation that stated that depending on the pressures and selection substantial adaptations can occur within 20 generations (think dog breeding). How that relates to genome is a bit different though.

[Devon Ryan]

Geographic spread is also going to be a pivotal factor here. If a species is geographically contained then the time needed for fixation is lowered.

[Michael]

I think if you could rephrase your question in the direction of molecular timing of the duplication event, this question would suddenly turn computational. Something along the lines, "can I calculate the molecular timing of a recent gene duplication event based on differences in DNA sequence?" You say there is nearly 100% sequence identity, what is it exactly? What would it mean if it was only e.g. 95% sequence similarity. I guess one should be able to calculate a confidence interval given an estimate of the mutation rate per nucleotide per generation?