Just found something additional. My wife's DNA matches that of someone whose ancestry is to Martha Brown, who was one of the children in the William and Eliza Brown family who appear in the Lamar County 1860 census with a number of children - some of whom appear to be from Eliza's previous marriage to a Davis (including Martha) and some who appear out of birth order (including a Mary A). Of course, I can't prove this genetic match belongs to this family (since AncestryDNA does not provide segment data) but the coincidence is pretty enormous since this is one of about 3 families I was considering as candidates for Mary Ann Brown. I will need to look for a Brown or a Davis who died and widowed an Eliza around 1850. Looks like I have a solid lead!
Don