On 2-adic orders of some binomial sums (Q710504)

The Catalan triangle \[ \left(\frac{k}{n}\binom{2n}{n-k}\right)_{n\geq k\geq 1} \] was introduced by Shapiro in 1976. He proved a formula for the \(n\)-th row: \[ \sum_{k=1}^n \frac{k}{n} \binom{2n}{n-k} = \frac{1}{2} \binom{2n}{n}, \] that is, \[ \sum_{k=1}^n \binom{2n}{n-k} k = \frac{n}{2} \binom{2n}{n}. \] Recently, variations of the Catalan triangle, such as \[ \binom{2n}{n-k}k^m, \] where \(m\) is a given positive integer, have been considered that extended above sum formula. \textit{V. J. W. Guo} and \textit{J. Zeng} [J. Number Theory 130, No. 1, 172--186 (2010; Zbl 1185.05006)] proved that, for \(m\) odd, \[ \sum_{k=1}^n \binom{ 2n} {n-k} k^m = \frac{n^2}{2}\binom{ 2n} {n} A, \] for an odd integer \(A\). They also conjectured that, for \(m\) even, \[ \sum_{k=1}^n \binom {2n} {n-k} k^m \] is divisible by \(2^{2n-\min\{\alpha(n),\alpha(m/2)\}-1},\) where \(\alpha(n)\) is the number of 1s in the binary expansion of \(n\). The current paper verifies this conjecture.