Those are quite nasty (by that I mean laborious), but they all seem solvable using things like Integration by parts, u-substitution (the "new" term for reversed chain rule.
First one: looking at it, it comprises of a composition and a product. Try u-substitution with either 1/x as u or e^{1/x} as u. I'm not bothered working it out that far. If u-substitution on its own fails, use IBP (int. by parts) and u substitution (but I would guess that would fail).
Second: what
>>4 said.
Third: Try using IBP where u (the one to be differentiated) is (tan(5x))^5 and dv (the one to be integrated) is 1. Think of the trick used to integrate ln(x).
Fourth: Similar to the second.
Fifth: Similar to the third.
These are educated guesses, I haven't bothered working them out, so don't expect my postulates to bear fruit instantly without some thought from you.