• Section S.2 of the Supplement states that "The fraction of trials whose maximum parsimony score is less than that of the of the original dataset is the empiricial p-value". The phrase "less than" should read "less than or equal to".
• The version of xscape used in the paper had a bug that allowed a parasite/gene node p on a host/species edge h to switch one of its children to the same host edge h. This effectively becomes a duplication. In regions of the cost landscape where switches are less expensive than duplications, xscape would choose these switches over duplications. This bug has been corrected in version 0.0.4 of xscape.
• Version 0.0.4 had a bug that caused xscape to swap transfer and loss costs when the ranges of these two costs were not the same. This has been corrected in version 0.0.5 of xscape.