Remember that (1+x)n is the sum of (n:k) × xk, by the binomial theorem. Differentiate this with respect to x, then substitute x = 1. The right side is the sum of k×(n:k), which is exactly what we want. The left side becomes n×2n-1.
If we are interested in the sum of k2×(n:k), take the second derivative and replace x with 1. This isn't quite right, because you get the sum of k×(k-1)×(n:k), but you can always ad n×2n-1 back in. This cancels the -k in k×(k-1). The answer is (n2+n)×2n-2. Note that similar formulas can be derived for the sum of k3×(n:k), and so on.