Lab update 2-15-08

In Lab: 

This week has been frustrating in several ways with regard to labwork.  First of all, I am attempting to optimize a particular procedure.  “Optimize” is code for “doing lots of experiments that are ultimately failures.”  While, in truth, you are actually moving forward because the optimization needs to be done and each parameter you tweak gets you closer to your goal of the (almost) perfect experiment (there is no such thing as the absolutely perfect experiment), it actually feels as though you are getting nowhere fast and have nothing to show for the 60 hours you just spent in lab.

The other reason this week was frustrating was because I ran smack into my own ignorance.  We have this image processing software that we use for the images that we take with the camera on the microscope.  I had been just merrily using the software, following the steps written up by someone else, when I ran into a problem.  When you run into a problem, that’s when you find out how well you know the system.  The answer for me was, “Not very well at all.”

The process I wanted to do is called deconvolution.  Before this week, I knew that deconvolution “made your image look better.”  How this happened, I really didn’t know (except that it involved math).  This ignorance (while not ideal) was not a problem for me in the past because everything worked the way it should.  But this week I ran into an instance where I had a really crappy looking image and deconvolution made it look even worse.  Which was counter to my thinking of what deconvolution was all about.

I now know that deconvolution is a mathematical way of assigning light back to its original source.  Think of a street light.  Normally, when you are looking at a street light, you see a halo of light surrounding it (particularly if it is foggy out).  The same is true for the fluorescent molecules that I am looking at in my sample.  However, I need to see exactly where the object is.  So now, think of two street lights really close together.  If they are close enough, their halos of light may become combined to look like one halo.  Now, if it was so foggy that you couldn’t actually see the street lights themselves, just the halo, you might think there is only one street light there when in fact there are two.  In my images, I really need to be able to tell if I am looking at one object or two objects really close together.  Deconvolution uses mathematical algorithms to get rid of the halo and only let you see the light that’s actually on the object.  It’s like getting rid of the halo on the street light such that you only see the light that is within the street light itself.

Did that make any sense at all?

Anyhoo, deconvolution is important in astronomy, too, for images coming from telescopes.  Because of this, my husband and I had several very long conversations discussing my image processing woes and how I might be able to resolve them (resolve–hahaha!).  Some couples may watch romantic movies on Valentine’s Day, we discuss image processing.

I was really hoping to be able to cross off some of the images on my list (to indicate completion) on the Paper Progress page, but alas it was not to be for this week.  Therefore, the highlight of my week was when a small rose plant arrived in lab for me for Valentine’s Day from my husband.

The Seminar Sock:

This week, the sock and I went to two thesis defenses (for people who are two years behind me in grad school years;  let’s not talk about how I feel about that).  The first one was from a student whose lab studies the mechanism of recombination during meiosis in yeast.  He had some really fantastic electron micrographs of DNA coated with his favorite protein.  He also had worked on an anti-cancer agent and had some interesting data about that.

The second thesis defense was from a student whose lab studies splicesome assembly and disassembly in yeast.  She had a huge amount of data.  Every time she came to a summary, I thought for sure that must be the last thing because it was so much work and then she went on to test another part of her model.  She’s going to have four papers when she leaves.  Amazing.  Her data was not as cool to look at as the first defense because it mostly consisted of a ton of RNA gels (which, all look about the same; I’m all about pretty pictures in science, this is probably why I’m in the lab that I’m in).  But, her method of systematically building her model was very elegant, as was the model itself.

I also have been working on the sock while taking images at the microscope.  I set up the scope to take a series of images.  I can’t leave the room while it’s doing so because then a ton of light would shine on the scope and I need it to be dark.  The computers are not hooked up to the internet, and it’s too dark to read well.  So, I knit.

Plans for next week:

I’m going to finish optimizing this technique if it kills me!  Then, I’m hopefully going to be able to take some images for my paper and cross off some of the figures on my list.

I’m also hoping to post on this blog a series of entries that I’ve been working on that explain immunofluorescence in layperson terms.  I’ve got to do a bit of tweaking on the posts over the weekend, but I should be able to put the first one up on Monday.  I’m also hoping to follow that up (during the next week) with some pictures of an immunofluorescence experiment in progress, culminating in pictures of the slides that I take on the scope.

Stay tuned!

As previously described

As a bench scientist, there is nothing I hate more about designing a protocol than going to the Materials and Methods section of a publication which used that technique and seeing this:

Because, then, having already taken the time to look up the first paper, you now have to look up the second paper in which sometimes–in an effort to make my brain explode or something–the procedure is cross-referenced yet again.   This can only be made more annoying by this fact:

The paper is from my lab.

Yes, folks, I looked in our book of Lab Methods with no success and I looked in a published book of protocols where this particular protocol was written by my advisor and the critical bit of information I needed was missing.  I then had to find a paper in which we used this procedure only to find the above explanation.  Finally, I tracked down the original paper which was fortunately not too difficult since it was from our lab (but I couldn’t find my copy so I had to get it from pubmed).

Needless to say, I was a little frustrated by the whole experience.  Fortunately, I now have the information I needed and I’m going to start the procedure right now.