<Also -- data mining and subgroup analysis are by and large the same thing >
They are not really the same. Usually the term data mining is when the data is looked at after the trial is done to identify any subgroups or stages of cancer where patient did better with the treatment than the general population. By looking hard enough one can always find some subgroup that did better, so this has limited if any predictive value for future trials.
Sub-group analyses that have some value are usually performed before the trial begins. These are usually designed to focus attention on groups thought likely to benefit from the treatment more than others based on some hypothesis of drug mechanism of action when specific mutations are present. In the case of SNTA and Ganetespib, one example is ALK+ patents in NSCLC.
Then just take the advice of what someone else here gave me: delay the purchase of snta until results are released....we will give upside but may reduce our risk and may give us some peace of mind. That is my strategy for now (subject to change like a million times before). Can't wait to results come out.
Where do you see this related to SNTA. It looked like reading what you posted that they had no success. Where as SNTA has had excellent results from trials. Actually I do not understand why I am even given your post the time to even reply. Wowwwwwwww.