Advanced analytic and machine learning methods have been proposed for safety signal generation and adverse event (AE) analyses. We employed two COTS software packages, Statistica and Spotfire Miner and a sponsor submission with a large number of subjects to evaluate the utility and usefulness of the software and underlying statistical methods. We present our findings for select adverse events of special interest, including an AE that appears in the product label as well as one that does not. We find that no single algorithm is "best" across all AEs and different performance metrics; however, a boosted trees algorithm does perform relatively well. We discuss the algorithms that evidenced better performance characteristics and those which did not appear to do as well.