RNA-Sequencing Vs. Microarrays
Or when, if ever, is using microarrays better than RNA-seq?
Gene expression Analysis is at the core of most of our projects at Genevia Technologies (more about our expression analyses here). There are two main assays yielding transcriptome-wide gene expression profiles of a tissue or culture, namely expression microarrays and next-generation RNA-sequencing.
A quick glance at Google Trends shows how microarrays, as a Google search topic, have steadily decreased in popularity since their heyday in the early 2000s. RNA-seq, having been in wide use for much of the past decade, still continues its rise in popularity.
The switch: the popularity of Google search topics “Microarray” and “RNA-Seq” since 2004 illustrates the technological shift in transcriptomic measurements. (Source: Google Trends)
This technological switch in gene expression profiling is no surprise, considering what RNA-seq enables:
- detecting novel, unannotated genes
- detecting sequence-level alterations (coding region mutations, gene fusions, A-to-I editing events)
- detecting alternative splicing events, even novel ones, unlike with standard microarrays
- broader dynamic range, adjustable by controlling the sequencing depth
- expression profiling of non-reference organisms (less straightforward but doable!)
- and so on.
In most cases, obtaining the expression profile of your sample is still a bit cheaper using microarrays instead of RNA-sequencing — the difference could be in the field of 50 to 100 EUR/USD per sample. However, the benefits of RNA-seq can easily outweigh the extra cost.
Yet microarrays have not disappeared. We regularly analyze microarray expression data for our customers (albeit not as much as RNA-seq). Researchers planning expression measurements still often ask our opinion on which technology to use.
The real question here is: when, if ever, does it make sense to use microarrays instead of RNA-sequencing?
Some times it does make sense. You might have a good reason to stick to microarrays, if 1) none of the issues listed above is critical, and 2) one of the following applies:
- you use a microarray-based diagnostic test with proven clinical utility,
- you have a large number of samples and cost is critical,
- you want to be able to compare the expression profiles directly with another microarray data set of the same array platform, or
- you have a running microarray workflow in-house (or with trusted partners) from sample collection to data analysis that you are happy with.
Lastly, one point from a bioinformatician’s viewpoint that you might want to consider: if you will have a researcher in your team use the data to learn bioinformatics, just remember that RNA-seq data analysis is definitely a more useful skill in modern (and future!) labs than microarray analysis.
The lists above should help you make the RNA-seq vs. microarray decision. However, it always pays to have a discussion with a bioinformatician or measurement provider to select the platform which optimally matches your biological material, research question and budget.