AsianScientist (Feb. 17, 2017) – I recently had a conversation with an undergraduate student who had joined our lab to work on his honors-year project. It felt like an interrogation and went something like this:
Student: How do you know what experiments to carry out?
Me: Well, that depends on the biological question you want to answer.
Student: So why do you use this particular cell line and knockdown this specific gene?
Me: The cell line is an in vitro model system, and the knockdown experiment lets you know if that gene affects a function in the model system.
Student: What about those ugly mice? What are they for?
Me: Those ugly mice are the descendants of a long line of heroes that have illuminated disease pathways generation after generation. They help validate whether our in vitro findings remain relevant in the context of whole-organism physiology.
Student: So when will you finish all the experiments and graduate?
Me: …
I think questions like these (except the final one) are the reason why honors-year projects are important. The research experience encourages students to think through the process of gathering scientific evidence, rather than simply memorize facts and findings written into biology textbooks.
Throughout the course of my PhD candidature, I’ve come to realize that the lesson of how to prove something has much greater value than what you’re trying to prove. In this month’s edition of Hacking a PhD, I thought I’d share some common approaches that biologists use to figure out how life works.
- Necessary and sufficient
For starters, the ‘necessary and sufficient’ approach to science-ing is old but gold. To explain its logic, we could use the analogy of a digital clock and a battery. If I take away the battery, the clock ceases to tell time. This suggests that the battery is necessary for the clock to function. If I replace the missing battery, the clock works again. This indicates that the battery is sufficient to restore the clock’s function.
How does this logic look like in the context of a biological experiment? A molecule known as a small interfering ribonucleic acid (siRNA) might be used to deplete a gene of interest in cells. We would then proceed to check the appearance and behavior of these gene-depleted cells. If changes are observed, we can infer that the gene of interest is necessary to maintain the cells in a default state, while the depletion of said gene results in an alternative cell state.
Next, we would reintroduce the gene (or its protein product) into the gene-depleted cells and observe whether they return to their default state. Should this ‘rescue’ be successful, one could claim that the gene of interest was sufficient to restore certain cellular functions. Such experiments are useful for establishing cause-and-effect relationships in biological systems.
- Time is of the essence
Biologists study cells and organisms which are, by their very nature, dynamic. Therefore, experiments to study biological systems should include the dimension of time. For example, when treating cells with an experimental drug, sample collection ought to be carried out at regular intervals so that the cellular response to the drug can be monitored over time. Data accumulated over multiple time-points can reveal patterns in cell or animal behavior that a study involving only a single end-point measurement would have overlooked.
However, time-point experiments exponentially increase the number of samples you’ll have to process. Two experimental conditions (treated cells vs. untreated cells) multiplied by five time-points in triplicate would mean that you’ll end up with 30 samples in total!
Another complaint about time-point experiments is the collection of samples at inconvenient intervals, such as once every eight hours. So you were in the lab at 8.30 a.m. to make the first sample collection? Your next two time-points are at 4.30 p.m. and 12.30 a.m. Is it any wonder this post containing the hashtag #overlyhonestmethods includes a confession about choosing experimental time points to avoid going to the lab at night or during the weekend?
- Different routes, same destination
From personal experience, journal reviewers really love to see a claim backed up by evidence obtained using at least two different methods. If you’ve looked at messenger RNA expression using quantitative real-time PCR, they’ll demand a western blot showing protein data. If you’ve performed immunostaining to show two proteins located in the same place within a cell, they’ll ask for a co-immunoprecipitation experiment. The use of more than one cell line for your in vitro experiments is also often recommended.
As scientific problems become increasingly complex and better tools are developed to obtain more conclusive evidence, the burden of proof on authors is likely to increase proportionally. Admittedly, these additional experiments take up more time, effort and resources. But here’s a tip: When preparing a manuscript, pay attention to the style and approach of studies that have already been published by the journal of your choice. This should allow you to gauge the type and amount of evidence required to convince the journal’s editors and reviewers.
With this, I wish you all the best in your research, and may the data be ever in your favor.
This article is from a monthly column called Hacking a PhD. Click here to see the other articles in this series.
———
Copyright: Asian Scientist Magazine; Photo: Shutterstock.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.
