deletions | additions       diff --git a/section_Results_General_Observations_Protein__.tex b/section_Results_General_Observations_Protein__.tex  index 43c49bd..0b0a59e 100644  --- a/section_Results_General_Observations_Protein__.tex  +++ b/section_Results_General_Observations_Protein__.tex 
...
\section*{Results}  Consistent with previous studies \cite{Tiikkainen_2009}, we find the MUV dataset to be a challenging target for shape-based screening with few targets demonstrating AUCs far from random performance. Overall FOMS either matched or exceeded the virtual screening performance of VAMS while retaining most of the benefits of pre-aligned molecules. Specifically, it is orders of magnitude faster than the optimized alignment of RDKit.  General Observations:  Protein targets with ligands that have many key interactions for their binding mode did not perform as favorably as targets with ligands that have a few key interactions. Targets with relatively few interactions may have the important features for protein inhibition covered by the SMARTS expression and fragment pre-alignment. These targets may easily reject compounds that did not match the SMARTS expression key for understanding the mode of inhibition.