X-ray crystallography

[solitude]

Lately I've become fascinated at the ability of biochemists/biophysicists to determine the structure of proteins. I really only know the basics of X-ray crystallography, and I know I could read about the actual technique elsewhere. But I'm wondering about the particulars of X-ray crystallography as a primary technique for a PhD? Are PhD's short/long? Is the process painstakingly boring or very exciting? What exactly does "solving a structure" entail? Is it common to get scooped when another lab solves a structure before you do?

I guess the gist of my post is, what do you guys think about doing a PhD in crystallography? Would you recommend it? I'm not applying until this summer, and certainly my eventual PhD lab will depend on the advisor and other factors, but I'm just curious because it looks so exciting in the abstract.

Thanks!

---

Members don't see this ad.

 

[RxnMan]

Lately I've become fascinated at the ability of biochemists/biophysicists to determine the structure of proteins. I really only know the basics of X-ray crystallography, and I know I could read about the actual technique elsewhere. But I'm wondering about the particulars of X-ray crystallography as a primary technique for a PhD? Are PhD's short/long? Is the process painstakingly boring or very exciting? What exactly does "solving a structure" entail? Is it common to get scooped when another lab solves a structure before you do?

I guess the gist of my post is, what do you guys think about doing a PhD in crystallography? Would you recommend it? I'm not applying until this summer, and certainly my eventual PhD lab will depend on the advisor and other factors, but I'm just curious because it looks so exciting in the abstract.

Thanks!

Crstallography is a technique - not a discipline, so you can't get a degree in it. You can shop around for a project that'll require you learning how it's done.

On a related note, while you may be fascinated by this technique now, don't choose your advisor solely on the project. Your advisor has the ability to make your work go faster or make it unbearable to continue in your program. First be confident that your advisor will get you out with a PhD in a reasonable amount of time and effort before you worry about the particulars of the project.

 

[solitude]

Crstallography is a technique - not a discipline, so you can't get a degree in it. You can shop around for a project that'll require you learning how it's done.

On a related note, while you may be fascinated by this technique now, don't choose your advisor solely on the project. Your advisor has the ability to make your work go faster or make it unbearable to continue in your program. First be confident that your advisor will get you out with a PhD in a reasonable amount of time and effort before you worry about the particulars of the project.

Right, but it seems to be the primary focus of a number of PhD projects (usually in a Structural Biology or Biochemistry department). What I'm wondering is what do people know about such projects and would they recommend them, assuming that the advisor is supportive, etc? As I mentioned above, the topic of research will only be one factor in choosing a lab, I'm just trying to get some idea of what people know about PhD's that spend their time doing X-ray crystallography.

 

[Circumflex]

I did the PhD portion of my training in a Biochemistry Department and knew a couple of graduate students who used crystallography as their main technique. They were miserable. Those projects tend to be very long and unpredictable. It is very cool if you enjoy physics and math - although not as directly relevant to medical issues or translatable research as other prjects would be. From what I understand, those students tend to spend more time earning their PhDs than others. Beware! You can always collaborate with a crystallographer and still learn about the technique.

 

[antinomian]

It depends if you want to push the boundaries of crystallographic technique, or use it as a tool to answer circumscribed questions. For example, Rockefeller's Rod MacKinnon has had to push the boundaries of the technique to solve questions about channel dynamics (membrane imbedded proteins etc...), while a guy here*, Eric Stebbins, regularly and quickly solves structures that are related to bacterial virulence factors to answer questions about pathogenesis. A guy like Nikola Pavletich at Sloan Kettering does a bit of both in working on cancer related pathways.

The point is that one can pursue crystallography as an end point, which is more difficult by the very nature of tackling hard questions/establishing new methodologies, or as a tool to look at various processes. The latter is not trivial, but much easier these days than it use to be. The difficulty of solving a particular structure is often (but not always) somewhat apparent from the outset.

In short, there is definitely a place for the technique in medical research depending on the way it is used.

*I'm only using people around here because i'm familiar with them - though Greg Verdine from Harvard's chemical/biology dept recently came and spoke about his work using creative structural analysis to answer mechanistic questions about DNA repair - excellent stuff, as well as his work on "undruggable targets"

 

[deadsea]

X-ray crystallography is more like an act than a science. You would never know when you can obstain the good crystal for diffraction. Physics and math are not essential. If the crystal is good, the rest are just automatic.

In the future, X-ray crystallography may become a comman technique like microscope. It is good to do a PhD in X-ray crystallography if you don't know which field you want to join now. Because you can join any field you like later.

I am receive the PhD training in a Structural Biology programme. That is my 2 cents.

 

[SpeakLittleB]

I agree, the crystallization part is very unpredictable. I've heard that graduate students generally get to wondering... what happens if this protein never crystallizes...

There are many standard programs for solving structures now, and it's relatively quick after you get the crystal. But making advancements requires a lot of technical knowledge. It seems like this is what you're interested though, the solving part, rather than the crystal part. So perhaps if you really want to be involved in this process, you'd be interested.

In general, though, i think it won't be shockingly exciting. Perhaps when you first get the crystal... but then it'll always be "ok, so that's the structure" which doesn't provide that much excitement unless you study it for some functional use.

 

[solitude]

OK so the gist that I'm getting is that you spend most of the time just trying to get your damn protein to crystallize, and after that it's just computer algorithms? That sounds pretty lame. It also sounds like it can be pretty easy to get some pubs if you're working on soluble proteins but much more difficult if membrane proteins (e.g. Mackinnon and the K+ channels).

Alright, well thanks for the info. I guess I will have to get more information and maybe do a rotation before falling in/out of love with that scientific discipline.

I agree, the crystallization part is very unpredictable. I've heard that graduate students generally get to wondering... what happens if this protein never crystallizes...

There are many standard programs for solving structures now, and it's relatively quick after you get the crystal. But making advancements requires a lot of technical knowledge. It seems like this is what you're interested though, the solving part, rather than the crystal part. So perhaps if you really want to be involved in this process, you'd be interested.

In general, though, i think it won't be shockingly exciting. Perhaps when you first get the crystal... but then it'll always be "ok, so that's the structure" which doesn't provide that much excitement unless you study it for some functional use.

 

[mercaptovizadeh]

OK so the gist that I'm getting is that you spend most of the time just trying to get your damn protein to crystallize, and after that it's just computer algorithms? That sounds pretty lame. It also sounds like it can be pretty easy to get some pubs if you're working on soluble proteins but much more difficult if membrane proteins (e.g. Mackinnon and the K+ channels).

Alright, well thanks for the info. I guess I will have to get more information and maybe do a rotation before falling in/out of love with that scientific discipline.

Try immunology. I'm a fan.

Regarding the X-ray crystallography, yes it is unpredictable, but it seems to me that they publish in excellent papers (Science and Nature quite often for an important protein). But yes, it is very haphazard, and the kinetics may be quite slow (I have heard of crystals being grown for a whole year?)

 

[solitude]

Try immunology. I'm a fan.

Regarding the X-ray crystallography, yes it is unpredictable, but it seems to me that they publish in excellent papers (Science and Nature quite often for an important protein). But yes, it is very haphazard, and the kinetics may be quite slow (I have heard of crystals being grown for a whole year?)

I really love molecular genetics, but immunology does look interesting. I'm actually taking an immunology grad course right now and I like it, but the grad students whine about having to work with mice, which can make the PhD drag on. I don't think there are many MSTP's in the immunology program here. I'm still a ways out from picking rotation labs, but maybe I will give an immunology lab a try when the time comes.

 

[cycologist]

I'm an M.D.(soon)/Ph.D.(done) who did Ph.D. work where x-ray crystallography was the major technique. I know people who have done Ph.D. work that only involved crystallography and the biochemical techniques necessary to make it happen. This, as others have implied, can drag on forever because there are rate-limiting steps at protein expression, crystallization, diffraction, data collection and structure solution. Any of these fail and it's no structure, no Ph.D.

Also, while it is easier to solve a structure than it used to be, it's still quite difficult (typically, for a totally new structure). However, just publishing the structure isn't good enough anymore unless it's a very interesting structure that stands on its own. Very interesting structures tend to be those that someone else has tried before and failed to solve -- you need to be brilliant or lucky or insanely hard working or all three to make these structures work. It is also possible to get entirely scooped by someone else, whereas people who use other techniques have results that most often don't entirely overlap with those scooping them. Feel lucky?

The good news is that it's lots of fun when it works, and given that it's easier than it used to be you can become a more well-rounded scientist by picking up other techniques to illustrate whatever story you're telling. A well-rounded paper with a structure and good biochemistry to go along with it is much more easy to get published.

Also note well that synchrotron radiation makes life that much better -- a lab that has access to synchrotron time is a bonus, but there are ways around that.

Also, I would note that while my math/physics skills are weak in an absolute sense, picking up the necessary background wasn't hard.

 

[juicybug]

I have done a few months of crystallography research, and I would say that's one of the most repetitive research you can ever do.If your research tries to crystallize an not well-studied protein what you will basically end up doing is this: purify protein, set up crystal tray, purify, set crystal tray.... looping again and again. I have also heard of people who spent the whole 7 years trying to get the structure. By the time they got the structure, another lab published the structure in nature and all their hard work was gone. ( That person got a PhD without any publication. That's insane!!!)
On the other hand, it can be pretty interesting if the type of crysallization you do only involves cryatallizing mutant protein with known structure. It takes much less time and the projects are much more flexible.
True, most other medical researchs also involve lots of repetition, but I simply can't stand doing research and not publishing paper in several years.
The buttom line is: talk to your advisors and see what type of crystallization research they do, how much time it takes to publish a paper in that lab...etc.