Top 5 tips for lab troubleshooting
By: Roche Life Science
Posted: October 09, 2015 | Everyday Essentials for Research
So your experiment didn't work. In fact, it brutally, and quite epically failed. You're not sure if the problem is the reagents, the design or some entirely different beast altogether. Now you've got to get to the bottom of the issue because your PI wants this experiment done ... like, yesterday. So now it's time to begin problem solving and figure out where things went wrong. Where do you start? Not to worry, in this article we will discuss our top five tips for troubleshooting your experiments in lab.
1. Retrace your steps: This sounds like what people tell you to do when you lose your keys or your wallet. It seems so simple, but when you recreate your actions, you can glean some additional insight into where things might have went wrong. This means checking the following:
Reagents (expiration dates, storage conditions, contamination).
Math (for any solutions or dilutions you may have made).
Protocol steps (did you forget a step or use the wrong solution).
Rarely the answer is obvious, but before moving onto more labor-intensive and time-consuming strategies, it's always best to be sure you didn't miss something. A kit that expired in 2005 or a dilution that was 100-fold off can really set you back in the lab down the road.
2. Know your protocol: Ideally, you will have thoroughly read and studied your protocol prior to doing the experiment. This will help make you aware of potential pitfalls and more difficult steps. Also, it's good to keep in mind that the more times you perform an experiment, the more confidence and comfort you will gain with the method, therefore, the better you will be at identifying potential headaches. Until then, when a new experiment is still quite unfamiliar, it's best to read and research. Look up similar protocols from various publications and websites to become familiar with variations in method. If you work from a kit, check the troubleshooting section in the back. While this information can be fairly basic, it can often point you in the right direction and help you figure out what to check next. If you use commercial reagents, read the package insert or check the website. You can often find FAQs and troubleshooting tips from the vendor website. If you know the vendor rep who works with your lab, you can even contact him or her directly.
3. Research your method: This is especially critical if you adapt a protocol from methods published in a paper. Be sure you have read the whole manuscript in detail and checked any references within the methods section where aspects of the protocol may have been developed in prior papers. These initial publications often contain more methodological detail than those later in a series. Sometimes, certain details may have been omitted, and you can consider emailing the corresponding author if needed to clarify essential points. Researchers usually are quite helpful when someone acknowledges and tries to adapt and utilize their work. Be sure to check with your PI before doing this, as they might have insight into the pros and cons of reaching out to potentially competing labs.
4. Get expert advice: The likelihood that someone else has experienced the same technical issue as you is quite high, so sometimes touching base with a more tenured researcher can make a huge difference and save you precious time and energy. He or she may be able to help shed light on your problem-solving approach and steer you to the right direction. This might be a fellow graduate student, post-doc or research associate from yours or another lab. If you don't have anyone with the right expertise to consult, you can always check online, as there are numerous forums where another person may have asked a similar question in past threads or where you can post your particular questions. Some of the most common websites for this are ResearchGate1 and Protocol Online2.
5. Carefully identify the variables and have the right controls: Lastly, it's time to test your problem-solving theories. To do this, you first need to have the right controls to know whether your experimental system is intact. This includes both a reliable positive and negative control because if either of these is not working properly, then something in your assay system is still an issue. Once you're confident the experimental design works properly, then it's time to methodically test your suspected variables in a stepwise manner. It is essential to change only one variable at a time because if you change multiple things at once, you won't be able to determine which modification was the necessary change. So make a list from most likely to least likely of the potential variables, and then proceed to repeat your experiment testing these modifications. Hopefully, your instincts were right, and it won't be long before your experiment is up and running with great results!
1 . http://www.researchgate.net/