I'm trying to reconstruct the ancestral sequence of an enhancer using IQ-TREE. I have the enhancer sequences for several organisms and their phylogenetic relationships (with branch lenghts included), coded in the following Newick tree:

((((((((((((((((Xiphophorus:0.163704,Poecilia:0.163704):0.30415,(Cyprinodon:0.40872,Fundulus:0.40872):0.059138):0.1919,Kryptolebias:0.659753):0.098759,Oryzias:0.758508):0.076102,Pundamilia:0.834611):0.011208,(Parambassis:0.82257,Acanthochromis:0.82257):0.023255):0.066068,(Scophthalmus:0.796596,((Mastacembelus:0.59803,Monopterus:0.59803):0.082706,(Betta:0.39546,Anabas:0.39546):0.28527):0.11586):0.11529):0.023867,(Gasterosteus:0.791517,Labrus:0.791517):0.14424):0.103656,Periophthalmus:1.04121):0.336975,Gadus:1.37639):0.433232,(Esox:0.82349,Hucho:0.82349):0.98612):0.38432,(Clupea:1.9049,(Danio:1.45489,(Electrophorus:1.2594,(Pygocentrus:0.85253,Astyanax:0.85253):0.40687):0.19548):0.45004):0.289):0.22731,(Scleropages:1.8936,Paramormyrops:1.8936):0.52768):0.073013,Anguilla:2.49425):0.53182,Lepisosteus:3.02611):0.55924,Acipenser:3.5853):0.82294;

The problem I'm having is that the node I'm most interested in is the first one (common ancestor of Acipenser with every other organism in the tree), which is not reconstructed by IQ-TREE. It reconstructs the ancestral sequence of each node starting from the second one (common ancestor of Lepisosteus and every other organism) onwards.

I've thought of a solution, although I'm not quite sure how good of a solution it is. I'm thinking of making up a "mock" outgroup, in which I've made up the sequence completely, but leaving the binding sites of the sequence (the parts I'm most interested in) as unknowns (coded as "N" in the sequence).

Would this be a good solution? If not, is there a better solution for this problem?