But if you want to hit a home run (ie., Cell, Science, Nature) then for most people that means 90 hour work-weeks.
Others may disagree with me, but I think getting published in those kinds of journals depends much less on your individual efforts and much more on how politically connected your PI is, what else your paper is competing with that month/week in that journal's submission, how carefully your reviewers are chosen/assigned, and how those reviews are handled. If you are on a strong project and get lucky, you get those papers. But I think on average, maybe with the exception of Cell, the
technical requirements for getting a paper into Nature or Science are less than into a leading "field" journal; the latter can have much more nitpicky reviews requiring actual work rather than creative wordplay. Sometimes we see the downside of that reality with big-time retractions from N&S. But I digress.
I think a lot of this is effective time management. Unless your experiment forces you to be in lab at odd hours/certain times, I think 50-60 hours per week is sufficient IF you spend that time efficiently. Not surfing Facebook waiting for a 3 hour incubation, but really multitasking. During my PhD (big lab, 4 years and 6 papers, ranging from PNAS x 2 to very specialized journals), I rarely spent any weekends in lab, but worked 10-12 hr days during the week, not taking significant breaks other than for lunch/coffee. I tried to make sure that my time was always occupied moving things forward, and for the most part that strategy worked out. Thus for me, it's less hours spent but how effectively you spend them. My PI also favored this strategy; we were only expected to be "around" in the late morning/early afternoon.
In general, your time during grad school is more flexible than during most of med school. Depending on the atmosphere set by your PI, vacations are easier to take and important events can be attended. I felt the time pressure with the combined program and worked hard, but burnout can easily set in in the face of months of no progress. Hence the late start times/short hours described by others, especially if you don't have this time pressure and your peers take 6-7 years to graduate. Unfortunately, and this is the downside to PhD life, your project won't happen unless you come in and actively work to make it happen. In med school, on the other hand, one
can get away with passive hoop-jumping depending on one's goals afterwards (i.e. not derm or super high-end IM residency, etc.).