Hi, Marc, it's Steven. Thanks for the question. So, obviously, as Hervé alluded upfront, I mean, this remains of critical importance to us and of major focus to us. And thankfully, despite the pandemic remains on track as well. In terms of this year's execution, there are three aspects beyond everything else going on in LIMBER, one is to get the monotherapy safety components of ALK2 and the BET program done and moved to the combinations of RUX plus ALK2 and RUX plus BET. And then as you alluded to, is start the pivotal Phase 3 of RUX plus PI3-kinase delta. They are slightly different. We don’t have a biomarker selection per se, but they're different intents behind the program. So if you look at ALK2, the combination there, it's important to both focus on enhancing efficacy, which we expect. But also safety in terms of anemia there and hopefully ameliorate the anemia through the hepcidin inhibition , which then would translate to being able to maintain ruxolitinib dosing and hopefully also enhance efficacy. So it's a dual play in that aspect. In terms of BET BRD inhibition, obviously, we've watched the consolidation data evolve both in the first-line and later settings. And if some maturity of that data needed, and we have to make sure there are apple-apple comparisons as much as possible in terms of patients treated, but the idea is to do the same thing there in terms of enhance the efficacy of ruxolitinib in those settings, particularly in second line and potentially in first line as well. And obviously, we'll just see where the data leads us in terms of getting that combination done as efficiently as possible. For delta, we very carefully built that data set. We've done different experiments with scheduling and dosing, we've now come down to the constant dosing being the way to go in terms of delta. In terms of the magnitude of the dose, we think we've weaved the therapeutic ratio correctly in terms of getting the efficacy we need from delta inhibition, but not the enhanced toxicity from it, and that's why we're initiating the Phase 3 studies. The bars may well end up being somewhat different in first line versus later lines. I think as we speak today, the first-line bar in terms of spleen volume response, which we set with ruxolitinib years ago remains a 35% measured reduction plus associated symptom reduction. In later line settings, we'll see as we work with regulators, whether a lower bar may be acceptable, for example something like a 20% screen volume reduction with associated clinical benefit in other endpoints like symptoms, and that remains to be worked out with regulators. So the programs are full steam ahead as much as possible given COVID-19, and we're confident that they are in a good place and then have slightly different intents to all three programs.