Tuesday, August 4, 2009

I think my head is bleeding

Fruit TartImage by Zeetz Jones via Flickr

... from banging it against the friggin' wall the past two days.

I pontificated this morning regarding the massive and often last-minute time sinks that I've been finding myself in largely due to the actions--or perhaps better said inactions and delayed actions--of Guru.

And it just doesn't stop. Most of the current head banging is grant associated but Guru-catalyzed. On top of the whole timeline issue I described earlier, he either can't make up his mind about what I should be doing or hasn't thought things through in the context of my project that's actually progressing.

I sent my specific aims--key word SPECIFIC--saying essentially that I want to study kiwis in an established in vitro model system. Then I want to study the action of kiwis in another in vitro system, where kiwis have not been studied but maybe more relevant with reference to biology. And then I want to examine (with important collaborators) the possible involvement of kiwis in fruit tarts in vivo. Guru replies that all the in vitro kiwi studies should be one aim with each system as a subaim and then to just mention the in vivo fruit tart model as a possible future direction. This seemed odd so I wrote to clarify. Good thing, as he had decided that we should indeed talk about fruit tarts. Great.

Now we're getting into what should be used as markers when establishing the role of kiwis in our in vivo model. I have my ideas based on broadly accepted markers. They are further downstream and could be influenced at multiple points by kiwis. But they are accepted markers of fruit tarts and are fairly easy to measure.

He wants to use pluots. He is very partial to pluots (perhaps because he is literally invested in pluots). Now, I realize that pluots are associated with fruit tarts in our model. Kiwis are associated with similar models but haven't been studied in our model. But pluots are observed early on, and we're not sure what kiwis are doing. I didn't argue the point with him because I'm new to (a) in vivo models in general and (b) the fruit tart model.

Yet the more I thought about the less sense it made. There is absolutely no evidence at this point for a connection between kiwis and pluots. NONE. Pluots are not the only contributor to the fruti tart. Not to mention pluots are very difficult to quantify. And this isn't just my reasoning. This conclusion was met by our resident fruit tart expert, as well. There's no way of knowing a priori if kiwis will affect pluots and even if they do what that means for our fruit tart. However, if we look at the more broadly accepted, functional markers, then we can determine if kiwis have an effect on fruit tarts--which is of interest to a broader community and to me even in the absence of pluot involvement. If so, then we can step back and examine if there is also an effect on pluots--which is of interest to Guru.

Not long ago, I actually described an experiment to determine if there was an in vitro correlation between kiwis and pluots. Guru responded that he would love it if they were, but they very well might not be, so we should focus on kiwis. Stupid postdoc planting seeds of hope...

What this last bit comes down to is that I was cowed by Guru yet again. And I'm pissed me off. Not so much at him right now (although it would be helpful to have conversations with actual explanations), but at myself. Because I have spent a lot of time going through the literature and talking to the ones doing the work in our lab. They don't have particularly high confidence in some of the previous unpublished pluot work. I know the problems with finding and quantifying pluots.

I feel like I'm starting grad school again. That I don't know anything, and surely Guru must know what he's talking about. But I know better on both counts. I may have a lot to learn about this field, but I know a lot about my project and the related work. I learned how to think critically and solve problems. Plus I realize that many PIs, especially established PIs with larger labs, have their brainwaves scattered over many things, so that they may not think through all the implications and difficulties of what they suggested. Yet when it came down to it, I let Guru batter through my confidence and run me down. Now I have to arm myself and batter my way back. I need to step up and take ownership of this project. It's my career, and I guess it's high time I started acting like it.

*Paramed commented that the pluots and kiwis were giving him a headache. Hopefully they won't affect anyone else too adversely. It was the best I could do at the moment. I'm happy to take suggestions for other non-science metaphors for research.


Comrade PhysioProf said...

One scientific phenotype to actively avoid is that of dancing around and around the key experiments that would test your pet hypothesis, but never homing in and lighting a motherfucking bonfire under your hypothesis. You should actively seek out the experiments that are most likely to disprove your favored hypothesis.

biochem belle said...

Oh, PhysioProf, you have some real gems.

I recall hearing a talk by Alan Fersht (one of the premier investigators in protein folding and function) in which he commented (with a dedicated slide written in large font) on this point.

To paraphrase: These days many experiments are designed with the intention of PROVING a hypothesis. This gets you into trouble because often you can never prove that your model is right, as there are usually other models that fit the data. He continued that when he was being trained, you stated a hypothesis and designed your experiments to FALSIFY said hypothesis. If the experiments didn't falsify it, you tried again.

I have to say though, your description of the process is much more vivid. I like it and shall have to file it away for future reference (with proper citation of course).