Problem Solving

One of the first things you learn as a researcher in a life science lab is just how unpredictable your results can be. Unlike physics research, in which things follow all sorts of defined rules (with the exception of quantum mechanics, but I'm sure they have their own rules), or history, in which you're browsing old (and constant) sources, life science labs depend a lot on living things, and they don't like to follow instructions.

Of course, our lab is better off than a lot of other labs, in that we generally just deal with DNA. We don't have to worry about picky bacteria or wilting plants or mice that don't respond the way you expect them to. But we still use "living" enzymes to do some of our work for us, and DNA can do weird things sometimes, too.

Sometimes, when our experiments mess up, we can blame ourselves. Maybe we didn't put in the right amount of something, or forgot to mix the samples well beforehand, or overmixed them into suds of broken, useless DNA (not that I have done that or anything...). And contamination is always a possibility - we are made of DNA, after all! However, no matter how perfectly you can follow instructions, sometimes things just don't work. And when that happens, you get to troubleshoot. That's often a large component of lab work, and pretty much the story of my week.

This week, we were working on getting a lot of our samples sequenced. Sequencing is expensive, and the more samples you can do at once, the better. On the way to sequencing, there are several steps involved:

1. Extraction - pulling the DNA out of your cell samples - in our case, cheek swabs or saliva samples. You can then measure the concentration of the DNA you collected.
2. Amplification - making more copies of the DNA you want using PCR, a method using enzymes that copy DNA and heat cycles. You can verify that it worked by running a gel of your samples.
3. Purification - getting rid of the DNA you don't want. Again, this uses an enzyme and heat

The extractions went fine last week! And then everything else didn't. Using two different instruments, we were getting way different readings for the amount of DNA we had! We had to re-measure several times before just deciding to go with the data we were getting. This was on Monday. We ran a gel of our PCR results (also on Monday), but nothing showed up on the gel. Generally, this means that PCR didn't work. So, on Tuesday, I ran another gel. And nothing! But on Wednesday, we measured the amount of DNA the samples had and everything - including the negative control that shouldn't have DNA - had DNA in it. Oy.

Once we figured out that the PCR process apparently didn't work, we redid it on Friday and ran another gel. Thankfully, we had good results on Friday afternoon. We still don't know why it didn't work the first time. But after 3 hours in lab today, all of our samples are prepped and ready to go for sequencing! So I'd call it a success.

Over all, this process can be completed in probably 2 days of lab work. Instead, we were working on it for a week and a half. And while things are not working, all you can do is think of what could have gone wrong - did I make a mistake? Should I redo the extraction/PCR/purification? Are there any old chemicals I'm using? It can really be a pain to try and figure out what went wrong.

And then sometimes, without you doing anything differently, it all works. And you get lucky! I'm certainly glad we did on Friday - better late than never!