Sunday, March 8, 2009

What's cooking and what's been koching?

I have two goals with this post: (a) to answer a question "why don't we have unzipping data yet" and (b) to explain some interesting simulation data we having related to an alternative splicing application of our shotgun DNA mapping technology.

What's been koching?

I have been getting asked the question a lot recently, in various forms and in various levels of directness: "how come you've had your new lab for two and a half years and you still don't have optical tweezers data?" This question comes about because of my mid-probationary tenure review, grant applications, collaborations, and just general behind-the-back chitter chatter. I have to stress that I find this question perfectly reasonable and totally valid. Most new professors get their lab up and running much more quickly than I have. Even I expected to be "up and running" sooner--but I am not disappointed. Rather, I am quite happy with where we are at, because my plan was to have the students involved in every aspect of our lab start-up, and to fundamentally understand optical tweezers, why we're building them, and what our impact is going to be. I've spent (or reserved) the majority of my start-up funds ($200K) on students. Those students are now poised to have a really great optical tweezers and a deep understanding of how we are going to make biological discoveries with this tool.

During my first two plus years as an assistant professor, I have failed at expressing my lab mentorship vision to my colleagues at UNM. I do think I've succeeded in privately communicating my vision to my department chair, mostly via my written annual updates and goals. But I've not communicated my plan to the vast majority of people who care about me, or at least care about me bringing in money. One of my graduate students, Andy, helped me realize yesterday this problem and I'm really happy that I now see it. Further discussion of this non-talent, I think I'll save for my other blog. My main point here is to just acknowledge that it's been my own failure to engage and communicate that has created this minor issue (only a major issue if I don't get my contract renewed, which is unlikely I think). My goal will be to much more aggressively communicate my vision to people when this issue comes up in face-to-face conversations. Maybe something like this:

"I know! I'm itching for the data to start coming out too. We've been much slower than average because my philosophy is that one of the missions of my lab is to mentor students. The students need to learn and need to be involved in the management of the construction of our instrumentation, the designing of our experiments, and even our grant writing. Because of that, we're not as far along in terms of data and publications as many people would have expected. On the other hand we are only a few weeks away from having a kick-ass optical tweezers system that the students have designed and constructed. And these students have a very solid understanding of what we're trying to achieve, and a self-driven motivation to achieve from this point onward. So, many good things are going to happen for us this year, and our rate of achievement over the next few years is going to be much, much higher than it would have been if I had simply built optical tweezers and commanded students to acquire data. Furthermore, because of their intense participation in our lab start-up, these students are going to be very strong in the next stages of their careers--particularly if they have to start-up their own lab."
I need to rehearse that. Sort of like my "30 seconds in the elevator with the CEO" kind of speech. Furthermore, when I'm asked this question by certain people, such as those that support all the extra BS and hoops that our graduate students have to jump through, I plan on going further to explain how those hoops and BS substantially impacted the progress of our research. What I have in mind here are the IGERT programs we have with ridiculous non-research requirements (including 4 extra courses) and our physics department's fixation on testing and retesting and retraining our graduate students on undergraduate quantum, classical, stat mech, and E&M.

The last thing I want to present in this section is a very clear example of the success of my student mentoring strategy: Andy, Larry, and Anthony (our three grad students) all applied for and were awarded individual research grants recently. These equipment grants were for $3,000 each, from UNM's graduate student association. Each of them drew up their own budget and individual research plan, and obviously they all succeeded in bringing in external funding for their research. They even went against my advice of combining their efforts to go for a single $3,000 award! I love that they proved me wrong, and I love even more that they have succeeded in getting their research funded at such an early stage in their careers. Each of them are purchasing specialized parts to make our optical tweezers better and they have thus succeeded in a major part of research: getting the funding so you can do the things you know you need to do. I'll end this section with a big congratulations to our grad students!

What's cooking?

Optical tweezers

We have a lot of things cooking right now, and like I said above, I think 2009 is going to be a really big year for our lab. Of course, optical tweezers construction has been a major focus. Anthony, Larry, and Linh have been working on this for over a year and achieved a number of milestones. Recently, Andy Maloney has brought his amazing optics and design skills to the project. He's also making a transition to open science, and has started publishing a lot of his work on OpenWetWare, Google 3D Warehouse, and elsewhere. Just one cool example is the current condenser / detection optics design, you can see a snapshot of Andy's sketchup design here.

Molecular biology and yeast genetics

In addition to constructing optical tweezers and control software, we've had a lot of things going on. One major area has been with our collaborator, Mary Ann Osley, working to create site-specific chromatin tethers and other biological constructs. Diego Ramallo Pardo (now at Stanford biophysics graduate school) initiated this work for us and made a lot of progress towards our goal of creating "unzippable" native chromatin fragments from a specific locus in yeast. Diego succeeded in transfecting yeast with a PHO5 plasmid having a unique I-SceI restriction site and he also succeeded in making dig/biotin DNA tethers. Further, he worked on our protocol for creating the versatile unzipping construct. This is a construct I designed during my graduate work. The beauty of this construct is that it allows for unzipping of virtually any DNA fragment that has a sticky end (see the figure). The versatility and power of this design has been demonstrated by students in the Wang lab who used the design to study helicase (Dan Johnson, preliminary work by Richard Yeh), mismatch repair protein (Lucy Bai), and mononucleosomes (Alla Shundrovsky, now Michael Hall), and of course my own work demonstrating the ability to map and probe protein-DNA interactions by unzipping.

Before he left for graduate school, Diego taught another student, Brandon Beck, what he had learned, and Brandon succeeded in creating unzipping constructs from an XhoI-digested plasmid DNA. The torch has now been passed to Anthony, who also has a strong natural talent for molecular biology. Furthermore, he's added an Open Notebook Science dimension to the project. You can find his notebook on openwetware. Anthony is focusing on creating shotgun clones from XhoI-digested DNA. We will use this to demonstrate shotgun DNA mapping of yeast genomic DNA--because we can sequence the clones to validate our single-molecule technique.

Alternative splicing: very promising simulation results

I want to update you on one specific result that Larry achieved recently: demonstrating that single-molecule analysis of splicing by unzipping looks very promising. Like all of the above fascinating unzipping studies, this too will be enabled by the versatile unzipping construct and the ability to unzip any DNA fragment with a known sticky end. We have been thinking that structural genome mapping would be a very cool application of our shotgun DNA mapping technology. (See our preprint on Nature Precedings.) Mary Ann Osley, our collaborator on our chromatin mapping project, suggested a couple months ago that we simulate an inversion mutation, since these are tricky to detect with ensemble methods, since they are "balanced" mutations. We asked around on friendfeed and got advice to check out "inv16" for acute myeloid leukemia. Unfortunately, Larry and I didn't know enough about genetics to figure out what exactly inv16 involved, and thus we couldn't easily simulate it. However, at that time, I remembered some very intersting material I had seen Scott Ness present, related to alternative splicing. It occured to me that unzipping would be perfect for analyzing splice variants, because large deletions, insertions, or rearrangements are easily detected by single-molecule unzipping. So, Larry found a recent paper from the Ness lab, about alternative splicing of c-Myb, which is the fascinating topic I'd heard Scott Ness talk about. Larry was able to find the sequence for a couple splice variants (8 and 8-b) on Pub Med, and put them into his unzipping simulation LabVIEW application. You can see the results of this simulation in the figure. Clearly, the difference between the two splice variants is one or more exons being missing in the black (variant 8) versus the red (variant 8b) curves.

We showed this simulation to the lead author on the paper, "Johnny O" a postdoc in the Ness lab. Even without explaining what the hell we were doing, he immediately saw what we were talking about. (Man I wish he was reviewing our grants!) (Johnny O is an excellent scientist, and he even has that sweet nickname, so definitely he is destined for greatness.) Our plan is to wait until (a) we have our optical tweezers going and (b) Anthony has mastered the DNA ligation method for creating the unzipping construct. Once we get to that point, we're going to get in contact again with Johnny O and see if we can develop a collaboration with him and Ness to try some single-molecule typing of cDNA clones of splice variants (using variations of our shotgun DNA mapping algorithm). The fact that cDNA clones have known restriction sites should make this experiment quite easy to try out once we have (a) and (b) going. This will be yet another example of the versatility of the unzipping construct I developed. And the fact that splice variants have large structural changes makes it highly likely we can succeed in single-molecule typing of splice variants (supported by our simulation above). I'm in love with this idea so much that I'm thinking it should evolve into a specific aim for our upcoming R01 applications. I'll talk more about that on my research blog, which I tend to use more for discussion of funding and future research plans.

So, that's my update on what's cooking now -- hopefully I've conveyed enough for you to understand why I'm excited for our lab and for our students in the coming months!

No comments:

Post a Comment

Creative Commons License
This work is licensed under a Creative Commons Attribution-Share Alike 3.0 Unported License.