thoughts on mouse projects now?

[wturk]

---

Members do not see this ad.

I've seen a lot of threads in the past about how MD/PhDs should avoid PhD projects where you have to generate your mice from scratch due to time issues. Any thoughts on how much this is still a factor with CRISPR, especially if the lab has never done it? Not very familiar with the amount of troubleshooting it requires vs. the old way.

More generally, thoughts on doing projects that require a technique that the lab currently doesn't do? Like genomics techniques in a mostly non-genomics lab.

 

[eteshoe]

Generating mouse models may be interesting but it really isn't a hypothesis driven project - you either make it or you don't (CRISPR can help but it's better to be in a lab that has a proven record of generating mouse models). And it can take quite a bit of time to generate these models (a year or more) depending on how complex of a model you are after (if you're even able to successfully generate it). I would imagine that's why it's better to avoid these types of projects unless you have a lot of infrastructure in place in your lab to make the model generation efficient.

For genomics techniques (like some sort of mapping or NGS technique of some sort), they can be picked up on your own or if your institution has a core that can teach you the basics. I guess I'd also consider the genetic complexity of your genetic model (e.g. yeast vs mice or whatever) when deciding to embark on a technique for which you'd have no support or help from your lab mates. It can be very cool, but you have to be efficient on the dual degree pathway and not waste time doing projects that will take too long to get up and running.

Another option would be two get a co-PI with expertise in what you're interested in so you don't spend a bunch of time blindly trying to build up your project.

 

[OxToCA]

For your second question, I did so recently and more or less figured out that collaborating and/or becoming a visiting scholar -- if your university allows it -- are good options. To increase your chances, tailor your research to generating co-authorship between your own and the lab doing the technique you're interested in toward supporting a grant proposal. If what you're doing would only lead to co-authorship, that's good on its own, too.

I had to navigate some pretty tricky things this year (e.g. creating an IND-supporting, IRB approved protocol for a radiopharmaceutical at one university which would be clinically studied and possibly considered for an early investigator award at another university while I was attending a different university since mine didn't have its imaging center built, yet). Notice that the attached grant application asks about non-federal (NIH) funds (e.g. from industry and the host University), multiple study sites, letters of support from collaborators, clinical trials, etc. Universities offer and encourage training for this, e.g. "The program is centered on helping students convert original research ideas into IRB-approved protocols that can be submitted for grant funding" https://synergy.dartmouth.edu/advanced-certificate, but they're mostly open to their own faculty, including the one I referenced.

 

[nonamesleft]

More generally, thoughts on doing projects that require a technique that the lab currently doesn't do? Like genomics techniques in a mostly non-genomics lab.

I dont do mouse, but generally, projects involving crystallization trials or animal model generating are rough on the mdphd timeline.
In terms of using techniques the lab doesnt currently do, it depends on what your support within the department/collaborations are. For me, majority of my work has been with techniques outside the expertise within the lab. For some of it, we have an expert in the department, so i can go to him and bounce ideas or get help analyzing and troubleshooting. For the parts where I dont have ready access to an 'expert', ive gotten some advice that ended up really eating into my time and didnt resolve the problem.

 

[Electricity]

In terms of timeline, the key divide that I can tell in my school is computational vs non-computational, with whether or not the technique has been done before in the lab not being as big of a factor. With respect to Crispr it will probably depend on what phenotype you're studying. E.g. Development should be okay, while aging or behavior, probably much longer.

 

[Lucca]

In terms of timeline, the key divide that I can tell in my school is computational vs non-computational, with whether or not the technique has been done before in the lab not being as big of a factor. With respect to Crispr it will probably depend on what phenotype you're studying. E.g. Development should be okay, while aging or behavior, probably much longer.

Just curious but do you mean purely computational PhDs finish faster vs wet lab?

 

[Electricity]

Just curious but do you mean purely computational PhDs finish faster vs wet lab?

Yeah, although not just purely computational. People with even an element of computational research seem to go faster. I've noticed this at my school but that's a pretty small sample size and it might be different elsewhere.

 

[eteshoe]

The pure comp people at my school still end up taking about as long as the wet lab folks since the stigma (it seems) is to make sure they're not getting an "easy" phd (lol whatever that means).

I have wet and comp portions for my project so I'll probably end up taking longer with my luck lol.

 

[MSTP18]

Haven't had too many, but the more computational people in my program definitely tend to finish faster or at the same time with more impressive work. My PhD was entirely wet but my work prior to grad school was entirely dry. It's not any easier, but when a mistake just requires identifying the spot in your code that needs debugging, that's a lot faster than say waiting weeks for your mice to age, vaccinating them, waiting a few more days, then infecting them, then waiting a few days to sacrifice then a few more days to get data to analyze. Also, you can realistically get data at home or while traveling or in very short bursts. If I'm not in lab, I can't run experiments. Certain experiments can only get done if I have 4 uninterrupted hours. If I go away for a week over the holidays it could take me a week from when I got back before I had stuff ready to do a real experiment. Those kind of time sinks are less common in computational work. Basically in a dry lab it feels like you are the real time limiting factor and the faster/more you work the faster/more data you get, whereas in a wet lab it can often feel like it's the experiment that's holding you back and no matter how hard/fast you're working, you can't get the data any faster. Hell, I felt like I was able to get more stuff done than some of my friends because all my work was in cells and the bulk of my experiments only took 1-3 days from start to finish to get data to analyze so I could **** up an experiment at the beginning of the week and get good data from the redo by the end vs. some of my friends who would get set back by an entire month when something went wrong.