This is a thread for all purposes X-ray crystallography, all other structure-determination methods are allowed. I am entirely fascinated with the work, remarkable work, that it takes from heterologous expression of proteins and their purification all the way down to discovering how these peptides shape life. There's so much literature out there that really puts the toil into perspective, and I have heard many stories of graduate students singlehandedly destroying their experiments, I would like to see whether Veeky Forums has some veterans that would share their experiences.
Also, a question.
As a junior, I've applied for a few REUs and internships that would help engross me in that sort of environment, but I honestly do not see myself getting them in the near future - mainly because I do not have impressive credentials. I would still love to get involved in this work field before my undergraduate years are over, but I have no idea where to go with the aspiration. Anyone care to shed some light? I would like to design a plausible experiment, but I don't know where to go with it.
My advice is to join a lab ASAP. Internships and REUs are great to apply to, but you should also start emailing PIs at your undergrad institution (or any universities or medical schools nearby) for your senior year. This would be the best way to start. I'd be glad to offer more advice if you're interested...
I didn't start undergrad research until my senior year. (And, I didn't even apply to REUs or internships. In fact, I don't even think I had ever heard of an REU at the time.) That's not too much time, but I was able to convince my then-PI that I needed to do a senior thesis and two semesters would work.
Wrote the thesis. Graduated, Took off a few years after undergrad, had some short-term jobs in industry my first year out. Then switched to academia, worked as a lab tech for two years. Went back to grad school for biophysics. Passed qualifying exams last year. Research interests in structural and biophysical projects.
Asher Scott
This is all very well and nice, but I have already undertaken a research opportunity under a professor, and they are very understanding of my situation, thankfully. I keep feeling horrible at how I wasted so much time immersing myself in other literature than working on the actual project assigned to me. I don't know. I finally feel like I have some sort of direction in my life now that I'm starting to abandon ship - this is bad. I'm slowly starting to recuperate from that fall, so we'll see.
I have discussed an opportunity to create an experiment with a respected professor at my university, but they said I would have to wait until graduate school, and this irks me. I would honestly rather do an internship without pay if it gets me to do what I want.
So you see how stubborn I am.
Brandon Morris
Bump.
Brody Nguyen
bump
Michael Nelson
Hey, not an experienced cristallographer myself, but I'm very interested in the field! We're going to take a trip to the Paris cyclotron in a few weeks to check out how they do experiments firsthand. There's a master thesis open that is actually about figuring out the structure and trips to Hamburg are needed to do so. It's all very interesting and one of the most fundamental biological studies one can participate in imo. Also, don't underestimate NMR and EM, they're great additions to the field of structure solving!
Cooper Peterson
>Also, don't underestimate NMR and EM, they're great additions to the field of structure solving!
This.
Cameron Parker
Watch out for the computational methods too bros ;)
I studied quite a bit in Biotechnology and related fields when I was in college and X-ray Chrystallography seemed to be painted as an antiquated method for determining the structural components of molecules. Is it still the go-to method for elucidating the characteristics of let's say.. a brand new peptide that was purified from a halophile?
Eli Gray
>Anyhow, have a nice and civil discussion. You're not my real dad, I don't have to do what you say. And you better not punish me again or I will call the cops and tell them you raped me.
Jacob Morris
I got a masters doing protein expression in plants, but I've not done much crystallization except for doing some screens which ended up not yielding any crystals.
If no one's going to let you work in his lab as an undergrad, then just take 18 credit-hours so that you can graduate as quickly as possible so that you can apply for a master's degree. You'll still have to get good grades, but if you can manage that, then you should be able to enter a master's program at Generic State U. There you can spend a couple of years purifying proteins, and if you get good enough at it, you can either do a PhD or try to find a job.
If you do a master's, remember that the goal is not to do any ground breaking research, but just to produce nice thesis that acts as an excuse to familiarize yourself with the following techniques:
1) Cloning in order to generate your expression vector, and expressing the protein (so, designing and ordering primers, using restriction enzymes, making mutations, and transforming and culturing your expression host). 2) Purifying your protein (i.e., polyacrylamide gels, enzyme assays, and using an FPLC) 3) Doing crystal screens (I don't have much experience with that; I think they have robots that can help you do this now) 4) Actually getting crystal data and then presenting the data
TRY TO FIND A MASTER'S PROJECT IN WHICH YOU ARE JUST MAKING A SMALL MODIFICATION TO AN ESTABLISHED PROTOCOL. THIS WILL GIVE YOU THE BEST CHANCE OF SUCCESS. When you do a Master's degree, you're just trying to get familiarity with the equipment and the techniques, and you want to do this in two years or less. You're not trying to do earth-shattering research. For example, maybe your project will be to mutate an amino acid on some globular protein that has already been crystallized by the lab, and then to crystallize the mutant protein yourself.
Zachary Howard
You really shouldn't worry about time spend reading literature in other fields; as an undergrad (in your sophomore or junior year, more so), you're still sampling different sub-fields for, at least, your senior thesis. People change fields all the time -- going from undergrad to grad school; from grad school to your post-doc; or even as tenure-track professors. You're still an undergrad, just about to join a lab -- your scientific career is just starting.
This is a good news: they'll take you as a graduate student. It shouldn't irk you; it's the way things are. The fact is that there's not much independent research that undergrads can really do, or much scientific training that undergrads can really receive. In undergrad, you're focusing largely on coursework, to provide the conceptual and theoretical background requisite to proceed to doctoral study. (Starting grad school, another year or two of graduate coursework aids in providing the framework for graduate research.) The undergrad degree is not a research-driven, advanced degree in science. That's what a doctoral degree in the sciences is.
I wouldn't say crystallography is outdated, but in recent years advances in cryo-EM have allowed people to pursue structures that have eluded structure determination by crystallography.
Depends on the question that's trying to be answered, and the tractability of the peptide for structural and biophysical study. Without knowing anything about the system, I would guess that a mix of x-ray crystallography and solution-NMR would be used to characterize the system.
Leo Hughes
Crystallography has been around for a long time, but it is still widely used for determining the structures of proteins because it gives very high-resolution data. There are other techniques based on NMR, such as deuterium exchange to measure surface accessibility, or solution NMR to get the structures for small peptides, but these aren't as precise as crystallization. You can also use some biochemical assays (e.g., cross-linking a transcription factor to its target DNA in order to determine how it binds), but these don't really reveal 3-D structural information.
The problem with crystallography is that it is often very hard to get usable crystals, but automated screening facilities have been designed and it is a lot easier now than it used to be even 10 years ago.
Juan Foster
I feel like advances in EM and atomic force microscopy is the future in terms of x-ray chrystallography resolution.
Charles Clark
I think they're beginning to complement each other more, instead of one method alone taking over X-ray refraction.
Jaxson Robinson
>I think they're beginning to complement each other more
I agree with this. I think that advances in EM will make the "architecture" (read: 5 - 8 A resolution) of large, multiple sub-unit, protein-protein complexes amenable to medium-resolution structure determination. In combination with high-resolution structure determination of sub-units (read: domains, if not proteins) -- high resolution that is not accessible by EM -- this can provide comprehensive structure determination of challenging targets.
Logan Rodriguez
>read: 5 - 8 A resolution I thought EM could reach 2 A res.
Blake Baker
It's extremely difficult to get to such a high resolution via EM. Also that resolution will be only in few parts of the structure, the core of the protein. The highest resolution for a protein was 0.48 by the way (ebi.ac.uk/pdbe/entry/pdb/5d8v/protein/1 ).
The general problem with EM is that you need quite big proteins or protein complexes to get your structure. Which is why it complements X-Ray crystallography so well. There your structure determination gets harder and harder with the size of the protein or the complexity. Growing crystals of huge proteins, especially membrane proteins remains a big challenge. Even with the advent of new techniques like LCP.
Brody Cox
AFAIK the best resolution for cryo-EM was 3 Å, but the thing is that cry-EM works best on big, symmetric molecules. X-Ray crystallography will probably never be beaten when it comes to high resolution atomic structures, but of course this only applies IF you can actually crystallize it.
Julian Richardson
I forgot to mention that X-ray crystallography doesn't really capture the dynamics of proteins, which are so important. You've got to NMR or MD that shit.
Ethan Thomas
Not quite. Cryo-EM can break 3 A (maybe 2.8 or 2.9, if I recall), but that requires large, symmetrical proteins or complexes. Needless to say, the difference between 2 and 3 A is huge. EM can complement crystallography; nowadays, it's not uncommon to do EM on a complex for a reconstruction at mid-resolution, then "fit" crystal structures of proteins or domains.
Oh yeah, there are a number of directions you can do with NMR experiments.
Isaiah Diaz
>doesn't really capture the dynamics of proteins Is it still possible to deduce them from crystallographic still images?
Liam Lee
>crystallographic still images You have no idea how x-ray crystallography works, do you?
>AFAIK the best resolution for cryo-EM was 3 Å, but the thing is that cry-EM works best on big, symmetric molecules. >Not quite. Cryo-EM can break 3 A (maybe 2.8 or 2.9, if I recall), but that requires large, symmetrical proteins or complexes. Needless to say, the difference between 2 and 3 A is huge.
Very interesting paper, I will go through it later today.
That said, there are some issues that I see with using cryo-EM in drug discovery. (Perhaps this paper will change my opinion.) The first is resolution. For drug discovery, generally, the lowest resolution you can get away with is 2.1 - 2.2 A. Ideally, you're using 1.4 - 1.7 A structures of inhibitors in complex with target.
Second, I would guess that there would be difficulties with class averages and the reconstruction. A small molecule would not be visible on the raw micrograph, so I don't think it's unreasonable to suggest that classes of particles could be grouped that are heterogeneous. In other words, you'd have 2-D classes that have a mix of inhibitor-bound and of native (ligand-free) particles.
You could get around this if the small molecules induce a conformational change in the target that you could visualize on a raw micrograph, or if the small molecules are targeting a protein-protein interaction. (The proteins would have to be large enough to pick-up on the micrograph. Typically in drug discovery, domains or peptides are expressed because they are more amendable to crystallography, solution-NMR, and biochemical/biophysical assay development.)
Xavier Clark
you define the type of crystal lattice? if yes what kind of software do you use for this?
Justin Ramirez
I agree with your concern. But the same also applies to the crystallographic solution of a liganded structure. If the ligand does not enforce some kind of conformational change then the lattice also does not change. And if you soak your crystal with your inhibitor then maybe only half of the protein molecules will take up the inhibitor. This would then result in an occupancy of 0.5 in the crystal structure. So also a mixture of inhibitor-bound and of native (ligand-free) protein molecules.
Alexander Fisher
rate
Connor Stewart
chuckled
Dominic Howard
Kind of, you can see what part of the structure are more uncertain than others, and then assume that this is because of flexibility.
Cooper Murphy
It made me happy.
Why is it that whenever I see anyone talking about crystallography it's a bunch of biologists/biochemists talking about biomolecules? Now, I did my undergrad in biochem, but I'm doing MatSci now, and I promise you all that XRD and X-Ray Crystallography are still some of the most powerful tools available to ionic, metallic, and other crystaline structures. There's a lot of interesting solid-state physics that you can explore with X-rays.
Liam Myers
I guess for this thread it is like that because OP already started with biochemistry instead of a general question about X-Ray crystallography.
Another reason might also be that in the high impact journals crystal structures of inorganic or organic molecules are just side notes at best hidden in the supplementary material section while at the same time a crystal structure of mammalian receptor proteins or complex membrane proteins is already worth its own paper and a cover page. It is more interesting for the interested public to hear how LSD binds in a serotonin receptor rather than seeing the structure of some crazy rare earth complex.
Daniel Young
We can talk about other molecules, if that's your area of expertise and or interest.
John Richardson
I think that this is a fair point. But, I would argue that there's a key difference. In the crystallography data set, after the data is processed, you can see the ligand density from the difference map. So, if you have only partial occupancy, you'll be able to tell that during the refinement. On the other hand, with EM, you really wouldn't be able to discern whether the compound is in the binding pocket (unless, again, there's a conformational change you can see on the micrograph that the compound induces). So, if you're picking particles and generating 2-D class averages for your reconstruction, you don't really know if the compound is there or not. Therein lies the difference; with crystallography, if you don't have full occupancy, your difference map will show it. I guess the difference comes down to the data processing, which in turn boils down to inherent differences in the techniques.
Kayden Hughes
Hello Veeky Forums! My username is "feynman" and I'm creating a discord for physicians, nursing, premed, and life sciences majors. Non-life science majors are welcome too. Please help populate our stem channel! For people who don't know discord, it's a free chat platform that uses text and voice.
