The problems pointed out in this post have turned into a major issue. A clinical trial was designed to test cancer treatments recommended by the erroneous analysis. The trial was stopped (and restarted and stopped again) and now the National Cancer Institute is conducting an investigation.