You’re about to start that big project you’ve been dreaming of for years. You’ve identified a potential miracle compound and want to figure out how it affects gene expression. But how are you going to do it: with next gen sequencing or a microarray? Especially if you are new to this area of research, the choice can seem like a difficult one, so here is a handy guide to point you in the right direction. Although I focus on gene expression, many of the pros and cons are also relevant to other applications, such as DNA methylation or sequence analysis. Just answer these five simple questions to help determine the choice that’s right for you!
1. What are the people around you using?
If you are a newbie to these methods or are not the person who will be doing the data collection/analysis, it’s important to look around and gauge what kind of expertise is available. If you plan on learning either technique, you will want plenty of help as you master the steps and troubleshoot problems. Is your lab already set up for microarrays? If the answer is yes, then you should give microarrays some serious thought. Similarly, if you work with a biostatistician who only wants to analyze next gen data, it would probably be wise to give this approach serious consideration. Switching from the technique most familiar to your research group may be warranted, but given the financial and time investments required to set up and learn a new method, it is a decision that should be made carefully.
2. How confident are you in your data analysis abilities?
One big advantage of microarrays is that they have been around for decades, which means researchers have developed a lot of really useful data analysis tools. Are you thinking about how best to normalize your data, for example? Over the years, scientists have developed many normalization techniques, so you can scan the literature to identify the one that best fits your purposes. In addition, lots of user-friendly software has been created to help not-so-tech-savvy researchers analyze their results. Although in-depth knowledge of the statistical methods being used is always preferable, programs like JMP Genomics and Ingenuity’s Pathway Analysis software make analyzing gene expression microarray data seem like a breeze. As more and more people adopt next gen sequencing, resources to help analyze these data are also becoming more common. Currently, though, in terms of both the sheer quantity of data and the level of sophistication necessary to analyze it, next gen sequencing is likely to pose a bigger challenge to the average researcher.
3. How much a priori knowledge of the genome do you have?
Your answer to this question may make your decision for you. To create a microarray, you have to know quite a bit about an organism’s genome. Do you want to study the effects of your miracle compound in humans? Mice? Pigs? Then you can order a few microarrays with a click of the mouse and a valid credit card number. If you are focusing on the four-petal pawpaw or the desert bandicoot, however, chances are that you are out of luck in the microarray department. Thank goodness next gen sequencing is an option! Next gen data allow you to analyze things like novel transcripts, splice junctions, and noncoding RNAs without having to know where they exist beforehand. Even for a genome that is relatively well characterized, getting such high-resolution data free of a priori knowledge is quite a bonus.
4. Are you interested in transcripts likely to be expressed at very low or high levels?
Microarrays indicate relative rather than absolute expression levels. This becomes a problem when the transcripts you are interested in are highly expressed, because signal saturation can occur. That is, it gets difficult to differentiate between two very highly expressed transcripts, even when a significant difference may exist between the two. Conversely, it’s easy for the expression of low-abundance transcripts to get lost in the noise inherent to microarrays, so if you want to assess rare transcripts, next gen sequencing is the option for you. Not surprisingly, many researchers appreciate the absolute expression levels produced by next gen sequencing techniques: the dynamic range is orders of magnitude greater than that of microarrays.
5. How much money do you have?
Although next gen sequencing is getting more affordable every day, many researchers still choose microarrays when cost is the deciding factor. If you are studying hundreds or thousands of samples, for example, chances are that next gen sequencing is cost prohibitive. Before getting scared away by the cost of next gen sequencing, however, it’s worthwhile to explore the resources available at your institution. Some core labs offer significantly reduced prices or pilot grants to encourage researchers to adopt the new next gen technologies they’ve invested in. Do your homework, and you may be pleasantly surprised!
Microarray and next gen sequencing approaches can be combined
Above, I presented microarrays and next gen sequencing as an either/or choice. The truth is many labs combine data from these two sources, allowing the strengths of one to compensate for the weaknesses of another.
- Take the example of a group of researchers studying the non-biting midge Chironomus riparius. You’ve probably never heard of this tiny creature, but it is commonly used in ecotoxicity testing. Marinkovic et al. wanted to study gene expression in this organism but didn’t have the a priori knowledge of the genome they needed to use a microarray. Thus, they made the obvious choice for studying gene expression: next gen sequencing. They knew this method was simply too expensive to use routinely in ecotoxicity testing, though, so they used their initial next gen sequencing data to create a microarray and detailed transcriptomic resources that could be used in future studies. Nice!
- More commonly, researchers might choose to validate results obtained using one method with results from another. Targeted RNAseq, for example, might be a great way to validate gene expression differences detected using a microarray, particularly if you are interested in looking at so many samples/genes that setting up that number of RT-PCRs makes you want to bang your head on your lab bench.
Hopefully answering these questions has helped you zero in on the approach that makes most sense for you. You can find tons of valuable perspectives on the microarray vs. next gen sequencing question on the web, however, in case you’d like to dig a little deeper. Here, I’ve picked a few of my favorite resources to share.
3. Zhao S, Fung-Leung W-P, Bittner A, Ngo K, Liu X. (2014) Comparison of RNA-seq and microarray in transcriptome profiling of activated T cells Plos One 9 e78644.