Mapping by Admixture Linkage Disequilibrium (MALD) is an economical and powerful approach for the identification of genomic regions harboring disease susceptibility genes in recently admixed populations. We develop an information-theory based measure, called EMI (expected mutual information), that computes the impact of a set of markers on the ability to infer ancestry at each chromosomal location. We then present a simple and effective algorithm for the selection of panels that strives to maximize the EMI score. Finally, we demonstrate via well established simulation tools that our panels provide considerably more power and accuracy for inferring disease gene loci via the MALD method in comparison to previous methods.