Now what we learned is we develop compounds that could hit SGLT2 or hit both that pharmacologic inhibition did not create that side effect profile, and that was a very important observation that we made first in our mice and our knockout mice and then using our compound that allowed us to have a therapeutic window of the advantages of SGLT1 without the disadvantages of the theoretical concerns.So that created a unique space for us and for our compound LX 4211. It profiled better in our preclinical models than the SGLT2 selective compounds we had and it continues to perform better, and we think now is profiling better in humans in clinical trials. So that's the evolution, the genesis of us of being here with a unique dual inhibitor. So our development strategy then. First and foremost, of course in diabetes you have to establish your agent has very robust glycemic control. It's a competitive space. There are a lot of agents out there, a lot of well established still patented agents, but also very old agents that actually do provide some level of benefit. So you have to demonstrate benefit on top of those. And ideally with a new mechanism of action, something that offers a new benefit to patients. And I think that our dual mechanism does offer that. SGLT1 mechanism, as we've been the only one to really explore this so far in the clinic, also provides certain additional benefits in glycemic control, specifically potent postprandial or after meal time glycemic control. That spike in glucose after meals which is a very important and somewhat convey to a certain amount of -- ultimately morbidity of diabetes is that spike right after meals. SGLT1 on the GI tract allows us to blunt that and I'll show you some data we have to that effect.