deletions | additions       diff --git a/bacoso4.tex b/bacoso4.tex  index 5184ef1..b2c8cd3 100644  --- a/bacoso4.tex  +++ b/bacoso4.tex 
...
In conclusion, our proposed strategy for structural prediction is as follows: first perform USPEX runs with spin-polarized DFT to generate the list of structures occupying local minima in the energy landscape. Then, apply LDA+U to the resulting structures to reorder the total energies to determine the true ground state structure. This is more economical than running USPEX with hundreds of calls to LDA+U, since LDA+U is more expensive than LDA. [Ran, could you estimate the difference in running time between your proposed re-ordering algorithm vs. the naive USPEX LDA+U run?]  \emph{Global stability} -- We did not construct the The  convex hulland check for global thermodynamic stability as the material was known to exist. [Perhaps we should.]  Our proposed algorithm  forstructure prediction generates several novel questions. How does $U$ affect  the energy landscape? In particular, does $U$ simply shift the local minima relative to one another, or does it create and destroy minima? Additionally, when Ba-Co-S-O system  is $U$ necessary for correct reordering shown in Fig.~\ref{bacoso-hull}. We find BaCoSO lies 102~meV/atom above the hull, which provides an estimate  of the candidate energies? We expect that $U$ systematic error within current DFT implementations, as the compound  is only necessary for compounds containing atoms with partially-filled $d$ or $f$ shells or magnetic materials, and this hypothesis deserves known  to be investigated. exist.