We looked at things differently. ARQL had samples from only 72 out of 167 ph2 patients, 50 out of 72 for IHC. So, I am not taking those numbers literally. The question I wanted to answer was why squamous population performed poorly in ph2 trial, the partial answer from the trial was consistent with majority literature that majority of squamous patients had no cMet expression at all, and consistent with MetMab ph2. For ph3 trial, squamous is irrelevant because they are excluded - I am glad they did. Of course there is risk with non-squamous population for ph3 alone without cMET as biomaker. That is a risk a clinical program should be willing to take at that juncture. If I were them, I would have done the same thing rather than running another prospective designed ph2 based on cMet biomaker. If they hadn't excluded squamous population, I wouldn't bet on them at all. Since they did, I give them a fairly good chance to succeed.
Market Data 
Markets 
AI
