As a preamble to my reply: my PhD from UCLA (2002), officially in “Biochemistry & Molecular Biology” (that’s what appears on the parchment), was essentially in the field of Structural Biology (which I’ll sometimes call ‘StrucBio’ here). In particular, I worked in the area of macromolecular crystallography, in David Eisenberg’s lab, with my interests being in crystallography itself (applied and theoretical), biochemistry and biophysics, RNA-binding proteins, and eventually in molecular evolution (the archaeal RNA-binding proteins that I studied are homologous to splicing-related proteins). I subsequently pursued postdoctoral studies at UCSD in molecular dynamics simulations of DNA and DNA-binding proteins (with Andy McCammon), so I also gained some familiarity with the StrucBio and molecular biophysics ecosystem at UCSD, ca. 2003–07. (Given that, the postdoctoral experience often differs greatly from a PhD; they’re apples and oranges, and I’m considering only apples in this reply.)

I mention all of the above as background information and an advertising disclaimer for my own answer to this question: UCLA is a remarkable place for structural biology! (At least it was when I was there, and I don’t believe it’s fallen since then.) For what it’s worth, I also want to mention that I chose to pursue a PhD at UCLA over appealing offers from UCSD (Chem & Biochem), UC Berkeley (Chem), Johns Hopkins (‘Intercampus Program in Molecular Biophysics’, as it was known then), and MIT (Chem). Each of those places were (and are) phenomenal environments for StrucBio and closely allied areas such as Biochemistry, Biophysics, Bioinformatics and Computational Biology, and so on. In my case, I felt there was simply no competing with the caliber of UCLA’s biochemistry and structural biology-related programs, at least back when I started there (late 1996): It was the environment of Jim Bowie, Rob Clubb, Dick Dickerson, David Eisenberg, Juli Feigon, Todd Yeates, all of whom were (or went on to be) luminaries in their respective areas of structural biology (National Academy members, HHMI Investigators, and so on).

But, to really return to the above list of places: One might think of conventional wisdom as screaming ‘MIT’ or ‘Berkeley’, but I suggest bucking conventional wisdom and seriously considering a range of places. By “range of places”, I really do mean a broad range in every sense—in terms of reputation & rankings, in terms of geography, in terms of the number and diversity of the sub-fields of StrucBio that are represented in the particular program. What are the most critical factors to consider in your “range of places”? To me, people are the single most valuable and crucial feature of any place. The humans that make up a given StrucBio program—their expertise and experience, and the diversity of their academic backgrounds; their personalities; how you mesh with them—to me, these are by far the most critical determinants of your success (and happiness!), not the number of CryoEM scopes or types of X-ray detector or HPC clusters or plate-readers or whatever. (Please don’t get me wrong—an environment with sufficient resources [‘resources’ writ large] for the types of science that you’ll pursue in your PhD is key too; notwithstanding that, hardware & instrumentation & equipment are secondary to people [and, not to worry: at the stage of first-year grad student, you needn’t worry about all those things—the environment should be already in-place for you to succeed, and you can rest assured it will be in-place at any of the institutions mentioned in this post].) So, really, in considering StrucBio grad programs, I suggest that the two most important criteria you consider are (i) the collection of faculty (lab heads) who would be available to you as PhD advisers (including the folks in their labs!—especially senior people, but really running the gamut, all the way from undergrads, to grad students, to postdocs, senior scientists, and all the way to tenured faculty—again, humans are the most valuable resource), and (ii) the simple number of such labs/faculty who you could realistically see yourself working for at institution A, versus B, versus C. This latter point is key: If there are 6 people/labs you could envision working for at Univ of Blah, only 2 at U of Yada, and maybe 3 people at Matamata Research Institute, then—all other things aside—your decision can be greatly clarified! (At least that was my decision-making approach, years ago, and it’s what led me to chose UCLA over other places [well, that, and—to be truthful—California…!].)

Note that the usual caveats that apply to this type of question apply to all of my above rambling: There’s no ‘right’ answer, and any answer depends exquisitely on many factors:

1. exactly what does ‘best’ mean to you ? And, do you mean it subjectively? (e.g., by reputation), or more objectively? (e.g., by PhD-within-5-year types of stats), or both ?…
2. closely related (but not identical) to the foregoing—what values/parameters/criteria matter most to you, as an individual human ?(Do you want to try surfing? Enjoy the outdoors? Like big city or small college town? Etc…) Again, no path is ‘better’ or ‘worse’—it’s all up to you and what works for you, yourself; that’s why it’s difficult to provide a generally-applicable answer for these sorts of questions…
3. I presume you mean graduate (PhD-track) level, not undergrad (there aren’t really structural biology degree programs at the undergrad level, though there are a few biophysics ones (e.g., UC Berkeley, and I think Hopkins… probably elsewhere).
4. How tightly focused are your interests, at least at present? E.g., do you already know that you want to work in RNA structural biology? If so, then note that Yale and Boulder have traditionally been two meccas for that; I’d also throw places like UCSC and Berkeley in this mix (Doudna, Cate, Nogales, et al…).
5. Are you limited to just the U.S.? North America? (U Toronto’s great!), Internationally? National Univ of Singapore (NUS) has been a rising star… alongside classic places in Europe like Oxford and Cambridge, where really much of Structural Biology was born (MRC!… and also places like UCL & Birkbeck for crystallography in particular); ETH-Zurich is also phenomenal for StrucBio, as are all the EMBL and Max Planck Institute labs in Germany.
6. Do you mean universities only, or are you also considering independent research institutes (e.g., the Scripps, Wistars, FHCRC, etc. of the world)?
7. and, as a catch-all: there are likely many other factors that I’m missing here!…

Finally, as a parting note: The US News & World Report rankings list the usual suspects—MIT, Stanford, Harvard, Berkeley, UCSF+Yale (tie)—in that order, from #1 → 5, and I really wouldn’t dispute any of those as top-flight places for StrucBio (in those sorts of lists, I think UCLA often winds-up in the #10–15 range under the ‘Biochemistry’ heading). Another list includes CalTech in the top-5, which is an assessment that I’d also not dispute. There are also numerous (literally dozens) programs/institutions with less general-purpose name-recognition than the Stanford’s and Harvard’s of the world (i.e., not the ‘beauty-pageant schools’, as one colleague refers to them), and yet which offer top-rate StrucBio programs. Some that come to mind are: U Washington, Seattle; U Oregon; Vollum Institute/OHSU [Oregon]; U Buffalo (Hauptman-Woodward is right there); U Chicago; Michigan State Univ; U Wisconsin, Madison; Univ of Utah; Baylor College of Medicine; U Texas, Austin; Rutgers; Columbia, NYU, Rockefeller, and Memorial Sloan Kettering in the NYC area; Cornell; UNC, Chapel Hill; Duke Univ; Vanderbilt; Arizona State Univ… the U Pittsburgh/CMU ecosystem (especially for computational structural biology—an area that would also include UIUC, RPI, Georgia Tech, Wash. U St Louis, U Kansas, Boston Univ)…There really are many great places, and everything in this post is a limited subset of what’s come up in my sleep-deprived mind (can try to add more later; others, please chime-in!).

Good luck—grad school can be a most amazing time in life, and structural biology is a captivating area to study— Enjoy it!!!