When I say safely I mean with out the risk of inbreeding.
I’m thinking something like one guy per hundred women to spread the genes around as thinly as possible? or something like that. (Which doesn’t sound like a bad plan to me btw…snickers)
SHAKES, face it, when they say “not even if you were the last man on Earth” they mean it.
Ha ha ha ha ha …ehhhem …ha ha ha. 
Suspicious glance
Why you ask?, are you planing to do something?
I think I remember that the minimum was about 2000 people; but I don´t know the male/female ratio.
In any case, I pledge to your indulgence to include me into the group, master. 
Strictly speaking to avoid inbreeding you only need about a dozen individuals. Inbreeding is just a complication resulting from receiving two copies of a detrimental recessive. To avoid that at any significant rate you really only need to avoid three things: sibling-sibling couplings, parent-offspring couplings, and double-first-cousin couplings (double first cousin couplings are the joining of the offspring of first cousin couplings with the offspring of another first cousin coupling that are the siblings of the first pair).
Beyond those pairings there is very little risk of inbreeding. As a result you need very few individuals provided you can maintain a strict control on who has sex with whom. There may still be problems but the population will survive. More rapid growth would be achieved with larger gene pools, but more than 50 unrelated individuals is pointless if inbreeding is your only concern.
The next problem is the one QED addresses. It’s got nothing to do with individuals. It’s about preserving genetic diversity. However you don’t need to worry about that because you intend to repopulate. Your population will be increasing as facts as possible I assume. As such you don’t need to worry about this because each generation will have more individuals with more novel mutations. QED’s quote concerns sustainable population size. Yours doesn’t need to be sustained at that size.
The next problem is simply genetic ability to cope with stress. How many people you need for this depends on what stress you intend to expose the population to. It’s very hard to predict what population we might need to cope with unforeseeable stresses. Some diseases have wiped out entire populations of animals in the past so no number was high enough in those cases. But if we assume that any novel stress will ‘only’ kill 99% of the population and that 50 is the number required to epopultate from that base, then we could assume 5000 people as a bare minimum.
SHAKES: We must prevent a mineshaft gap!
Q.E.D.: Heck, 500 people is well under the population of a small town. I’m assuming that all of them are healthy, reproductive age people who will freely interbreed, and that none of them are from the same family? The article only mentions the ratio of males to females.
If 500 is all you need, some long-term colonization scenarios become much more plausible.
DAMMIT!
sniff I need a hug… 