Molecular biology voodoo: Embracing imperfect science experiments

What do you do when your experiments simply won’t work? That’s a pretty broad question, so let’s narrow things down and provide some context. In my experience, there are two main scenarios. The first is when a protocol that has always worked for you suddenly stops working. Scenario No. 2 is when you’re starting a new project, and despite the expert protocol that was left for you by a previous student (usually scribbled on a Post-it note or napkin), you simply cannot get the procedure to work. First off, these “failures” are important aspects of your scientific growth that you need to embrace. Secondly, it’s equally important to mention things that you should not do in these situations.

Let’s consider scenario No. 1, with the hypothetical situation that you’re given the task of cloning a particular gene, using a standard cloning vector with blue/white screening. You’ve performed this task countless times, and it shouldn’t take you longer than one week at most. Things begin promisingly, as you are easily amplifying your gene via pcr44974338_s. You move on to ligation and subsequently transform your product before you leave for the day, fully expecting to return the next morning with a nice mixture of blue and white colonies on your plate. But after removing your plate from the incubator, not only is there not a mixture of blue and white colonies, but your plate seems to actually be cleaner than when you left the previous night! What do you do (besides, of course, what everyone who has ever had a failed transformation has done, and hold the plate up toward the light, squinting as you try to determine if that errant air bubble is actually a rogue colony)?

You decide to play it cool because you know these things happen from time to time. And since you saved a portion of your transformation from yesterday, you plan to replate it before you leave today, confident that all will be better by tomorrow. Then, to your dismay, you get the same result—a miraculously clean bacterial plate. This is when it’s important to know what not to do.

Do not start pointing fingers, despite believing that you performed the experiment perfectly. The fact that everyone else in the lab is successfully cloning should make you realize that your lab mate didn’t contaminate the reagents. Go back to step one and start over. For whatever reason, this clears up the issue most of the time. This is when you need to accept that as much as we like to have things down to “an exact science,” when it comes to molecular biology, there is some voodoo involved at times. If it worked the second time and you can’t tell what you did differently, just move on and be glad. Think of it like a quarterback who throws an interception, or a basketball player who has an inexplicable bad shooting night (I like to consider myself the LeBron James of cloning), only to bounce back and win the next game.

What about scenario No. 2, when you’re trying to get a new protocol to work for the first time? What I’ve found works best when challenged with this is to go to the literature and find multiple examples of the technique in question. This is a great way to test yourself regarding your understanding of the protocol, as you’ll notice slight differences between papers. When my PhD advisor informed me that I’d be doing in situ hybridization experiments, I didn’t have a clue what that entailed. If you’re unfamiliar with in situ hybridization, it’s a technique used to determine the spatial and temporal expression of a gene of interest (typically using an RNA probe). It can be a very tedious procedure that takes three or more days to perform, and requires multiple reagents to be made, in addition to collecting the samples and synthesizing the probes. At this point, I was less than a year into grad school, so I was unfamiliar with the technique, to say the least. I initially approached the procedure head-on, using the protocol given to me—from a very reputable paper, nonetheless.

After a few months of minimal success, my frustration began to mount. What could I possibly be doing wrong? I felt I was following the protocol exactly. I checked all my calculations. The reagents were made properly. So I headed to the library and found a book called “In situ hybridization, a practical approach.” I was curious: How could someone fill an entire book on how to perform this procedure? Flipping through the pages, it seemed to have the same basic steps I’d been fruitlessly doing the last couple months. But upon further reading, I began to notice areas of the protocol that differed slightly from the one I had. After finishing the book, I had a short list of differences that I couldn’t explain. I took them to my advisor, and, one by one, he began eliminating them from the list. Most of the differences didn’t matter. However, there were a few things he honestly didn’t know. As a graduate student, when you’re pursuing something and your advisor is at a loss, you know you’re onto something.

With a renewed sense of direction, I returned to work. My approach changed, however. For each reagent used, I asked myself: “What’s its purpose? What would happen if we eliminated it from the protocol?” Doing a quick PubMed search (with only the goal of finding papers using this technique), I was able to compile another list: reagents that were mentioned in every paper, and reagents that seemed to be optional. I started eliminating some of the optional reagents, and low and behold, I started to get results! This is where the voodoo that is molecular biology comes into play. It’s not that one paper is wrong and another is right. Sometimes an experiment works, and when it does, it becomes part of the protocol. Like the quarterback who throws an interception but must push forward for success, the quarterback who throws a touchdown pass also moves on, leaving behind a seemingly perfectly diagramed play that certainly won’t always be successful. Sure, there are fundamentally important elements of the play’s design and execution, but some attributes are supplementary or even coincidental.   

Years later, I have a much different perspective of this scenario. Many lab experiments we perform have multiple intricate layers. Therefore, applying a one-size-fits-all approach simply won’t work, and if you do so, you’re setting yourself up for failure. Every year at Thanksgiving, we basically prepare the same meal. Yet how is it that some years the outcome is so different than others? Wouldn’t you think that you could follow the same recipe (protocol) each year and get the same results? The turkey my father-in-law made last year was out of this world, and it’s possible that all those stars may never align again. Still, I’m sure my father-in-law will try to top that turkey this year. But what if he doesn’t? Are we going to kick him out of the kitchen (lab)? What if he goes into a slump and the turkey quality continues to decline each year? I’m certainly glad my advisor didn’t kick me out of the kitchen—er, lab—when my in situ hybridizations weren’t out of this world. I finished my PhD, and now I’m a postdoc. I just started doing in situ hybridizations here in my new lab. And you know what? They aren’t looking as good as they did when I was finishing graduate school, but at least I know they’ll work—most of the time.

Quartzy is the world’s No. 1 lab management platform. We help scientists easily organize orders, manage inventory, and save money. We’re free and always will be. Visit or reach out at

Interested in writing for The Q? Send us an email!

Dr. Drew

Dr. Drew obtained his PhD in molecular biology while building a family of 5 with his lovely wife. Dr. Drew is currently expanding his horizons and working as a postdoctoral research fellow where he studies congenital heart defects. In his spare time he enjoys ball room dancing with his wife.

1 thought on “Molecular biology voodoo: Embracing imperfect science experiments”

Leave a Reply

Your email address will not be published. Required fields are marked *